Designing Blower Systems.

Blowers and fans generally provide air movement for industrial processes and for ventilation. The difference between the two devices is the system pressure they operate against and the method used to move the air. Blowers by comparison tend to operate at higher pressures than fans. 

When designing a system which utilises a blower, the engineer will typically need detailed knowledge of the application and its requirements. This means, an estimate of the blower design duty point, i.e. the flow rate required and the resistance to this flow rate arising from ductwork, bends, filters, fittings etc. This resistance is also known as system resistance, backpressure or head loss. Figure 1 provides an overview of the curve 

The blower capacity/performance curve represents the blower’s ability to deliver flow against backpressure, from zero backpressure (“open flow”) to completely blocked flow (“sealed”). Note, the single blower performance curve shown in Figure 1 would represent the maximum blower performance at a constant electrical input. 

The system resistance curve characterizes the pressure drop in the system for a given flow rate.

When both the blower performance curve and system resistance curve are overlaid on the one chart, the blower duty point can be established. The required performance can be achieved by a number of blower models however, each will exhibit a different efficiency. The ideal blower selection will be one which offers the desired performance at an optimum efficiency resulting in the least amount input power required. 

Blowers are often speed controllable which means in a system with dynamic operating conditions, the speed can be reduced to a desired operating point or increased if a different operating point is required in a dynamic system.  

Changes in blower performance based on speed changes is reasonably predictable using the fan laws and Table 1 describes the equations used to predict the change in flow rate, pressure and power as a result of changing the operating speed. The subscript 1 represents the initial existing condition and subscript 2 represents the revised calculated condition based on speed change. 

Table 1: Fan Laws Related to Speed Changes.

Static and Velocity Pressure

The energy of a fluid can essentially be viewed as having two types of pressure, static pressure and velocity pressure. Static pressure is the force per unit area exerted by the fluid in its surroundings and is independent of the fluid motion. Velocity pressure is the velocity head or kinetic energy content of the fluid. It is the pressure that would result from reducing the fluid velocity to zero, converting the velocity pressure to pressure head. The summation of the static pressure and velocity pressure is known as total pressure or stagnation pressure. 

In any flow system, there is likely to be elevation change (potential energy or elevation head) that also contribute to the energy content of the flow stream at a given point in the system. These changes in elevation head or potential energy for gas flow is often quite small can is sometimes ignored. However, when blowers are modelled in FluidFlow, these effects are taken into account to develop a more accurate reflection of the system operating conditions. 

Blower performance curves are typically plotted as static pressure rise against flow rate. The velocity pressure is typically not a useful energy quantity for overcoming resistance to flow. So, in general, the resistance to flow results in a loss of static pressure of the streamline and not to a loss of velocity pressure. 

Fluid Density

The density of the fluid being transported can have a significant effect on blower performance. Table 2 describes the equations used to predict the change in blower performance as a result of fluid density. 

Table 2: Fan Laws Related to Fluid Density Changes.

As fluid density is independent of the blower, it is important to specify the density of the fluid as part of any equipment selection or performance specification. When modelling blowers in FluidFlow, the fluid density is automatically traced throughout the system and the effects of density changes included in the solved system. 

Blowers are often used to move air however; they are also used to move explosive or non-explosive gas mixtures. In some cases, the gas mixture is considered to be an ideal gas. FluidFlow doesn’t make this simplifying assumption but solves for real gas conditions using an equation of state. 

Compressible flow systems which use blowers can be designed and modelled using FluidFlow software. The blower performance characteristic curve can be defined in the database using your preferred units for flow and pressure. When using volumetric flow rates, the units can be reference to either STP or NTP conditions. The unit’s environment can be customized to display the units you prefer to use, i.e. mass flow, volumetric flow, pressure in psi, kPa etc. 

Figure 2 below gives an example blower performance curve. This curve data is a based on a pressure blower unit by the New York Blower Company at a speed of 3500 RPM. 

Figure 2: Blower Manufacturer’s Performance Curve.

The performance data was then added to the FluidFlow software database. Figure 3 then gives an overview of the data-entry utilised in the software. 

A system design is developed as shown in Figure 4 and the blower performance evaluated. System static pressures, velocities, friction losses, flow rates etc can all be evaluated and the system performance optimised.  

In summary, when blower systems are modelled in FluidFlow, the software doesn’t make the simplifying assumption of the ideal gas equations and does take account of density changes as well as the effects from elevation changes in the system. Modelling the system also lets designers clearly establish the operating conditions of the connected system and consider dynamic operating conditions, i.e. changes in operating speed etc. 

FluidFlow v3.47


Bug Fixes
* Fixed a bug affecting pipe inlet results for pulp/paper calculations only where the viscosity at a pipe inlet was displayed incorrectly.

v3.46 (CUSTOMER RELEASE) – 13th March – 2020
* Network Login: User prompted to confirm before continuing.

v3.46 (MAINTENANCE RELEASE) – 6th August – 2019
Bug Fixes
* Composite Chart: EGL/HGL fixed.

Additionally, the help has been improved and updated to reflect any recent changes/additions.

Heat Transfer in Pipes

Heat transfer between the fluid and the pipe’s surroundings is an important aspect of piping system design and a computer program is often a useful tool to use to help understand the effects of heat transfer. The addition or removal of heat to/from the pipe flow stream can change the fluid viscosity, change the fluid phase-state, i.e. change from liquid to solid phase, cause cavitation, it can cause the surroundings to freeze or melt and in the case of gas flow, can cause significant changes to the fluid density.  

Pipe heat transfer can be quite a challenging task and can become increasingly difficult as layers such as insulation are added to the pipe. 

Heat transfer should be combined with fluid flow analysis to achieve an accurate estimation of the fluid physical properties related to fluid flow such as density, viscosity and in two-phase flow, liquid surface tension. Also, fluid flow thermodynamics such as vapor-to-liquid ratio which is an important factor in two-phase flow of gases and liquids is a function of the fluid temperature. 

Pipe heat transfer to or from the surroundings can occur by one or a combination of the following three known heat transfer processes, convection, conduction and radiation. 

Convection: This refers to the transfer of heat energy by movement of fluids. Convection heat transfer is a function of the fluid physical properties, fluid velocity and system geometry (pipe diameter). Pipe flow is normally subjected to simultaneous heat transfer by convection and conduction. Convection heat transfer can be divided into two types; free (natural) or forced. Free convection refers to heat transfer which is driven by buoyancy, i.e. it occurs due to the natural movement of fluid caused by a density difference induced by a temperature difference. Forced convection arises when the fluid is forced to move by an external source such as a fan. 

Conduction: This refers to the mechanism whereby an energy exchange occurs between the fluid and the pipe wall, insulation and soil (if the pipe is buried) due to direct contact.  

Radiation: Radiation heat transfer is the electromagnetic transfer of energy from a hot surface to a colder surface and represents heat transfer without a physical medium. Although radiation heat transfer is often minimal for short pipe lengths, it is an important aspect of heat transfer analysis of onshore oil and gas transmission lines. 

Pipe heat transfer calculations are often based on heat transfer rate per linear “M” or “ft” rather than the area of the length of pipe under consideration. In carrying out pipe heat transfer analysis, the various diameters of pipe insulation must also be considered, i.e. the inner area of the insulation is lower than the outer area and this difference must be considered in the analysis. 

Heat transfer in piping systems, in its simplest form, can be for a single uninsulated pipe. However, it is often the case that the pipe has a layer or several layers (insulation etc) which the complicates the analysis. Figure 1 gives an overview of a cross-section through a pipe with a number of concentric, cylindrical layers. 

Heat Transfer in Pipes


Q is the heat transfer rate (btu/hr).

k is the thermal conductivity of the material (btu/h ft F).

R is the radius (ft).

T is the temperature (F). 

L is the pipe linear length (ft).

Performing pipe heat transfer calculations allows designers to better understand system operating conditions and select the most suitable and economic thickness of pipe insulation required for a given application.

Below are a number of points which highlight some of the main benefits of insulating piping systems:

  • Insulation helps reduce and delay the heat transfer into or out of a process which in turn helps maintain a stable fluid temperature, allow chemical reactions to proceed normally and safely. 
  • Uninsulated pipes permit the uncontrollable release of thermal energy to atmosphere which wastes millions of dollars. 
  • Insulation offers protection against freezing. If piping systems such as steam condensate, aqueous solutions, fire water systems were uninsulated in many cases, the would be a likelihood of freezing thereby preventing the operation of the system. Freezing can also cause failure of components as a result of water expansion when it freezes. 
  • Insulation prevents moisture condensing from the air onto cold pipe surfaces which can cause damage to surrounding materials and equipment. 
  • Insulation also offers protection against burns. In industries such as petrochemical, chemical process etc, pipelines can operate at very high temperatures and in locations where personnel need access and need to work safely on a daily basis. Providing insulation to pipes and fittings in these hazardous areas therefore gives protection to the employees allowing them to work safely and effectively in these zones. 
  • Since insulation retards uncontrolled thermal energy transfer from pipes, it reduces wasteful heat loss which has a knock-on effect of reducing wasteful energy consumption and associated emissions. 

Insulation of piping systems serves an important function in the operation of oil and gas facilities, petrochemical processes, chemical processes, oil refining plant as well as chilled water systems, district heating system and refrigeration plant, to name a few. In some industries, refrigeration systems can account for a significant proportion of overall site energy costs. The correct selection and use of insulation on these piping systems can significantly improve system efficiency, with thermal performance and condensation prevention being the key considerations.

Economic Thickness of Insulation

When choosing insulation thickness for piping systems it is important to note that there is a limit beyond which it is not prudent to increase the insulation thickness. Increasing the thickness of insulation will reduce the heat loss from the pipe and thus give a saving in terms of operating costs. The cost of insulation increases with increase in thickness and there is an optimum thickness when a further increase does not save sufficient heat to justify the cost. In general, the smaller the pipe, the smaller the insulation required. So, a well-designed and optimised piping system will help with the selection of the economics of insulation thickness required.

There are various heat transfer functions available in FluidFlow. As well as performing heat transfer analysis at heat exchangers etc, pipe heat transfer analysis can be completed using a number of different approaches. Firstly, above-ground or buried-pipe heat transfer calculation can be considered. When modelling above-ground scenarios, you can perform heat transfer calculations for uninsulated or insulated pipes. You can set a fixed temperature change across the pipe or a fixed heat energy transfer across the pipe. 



District Heating Systems

District Heating Systems provide a method of delivering thermal energy to buildings (homes and businesses) in the form of hot water through a distribution network of highly insulated pipelines. In this way, heat rather than fuel is delivered to buildings which are connected to the district heating system. As such, the buildings don’t require independent and dedicated heating plant meaning there is also no requirement for gas or oil deliveries to the buildings. A heat exchanger at each building transfers the heat energy from the district heating system network to the building’s own water-based heating system which in turn supplies heat for both space heating and domestic hot water requirements. 

District heating systems have the ability to utilize and distribute heat from many different sources, including heat sources that are typically seen as a by-product and can usually go to waste, such as waste heat from electricity production and industry. To put that in some context, it has been estimated that there is enough heat currently going to waste from electricity production and large industries in Europe to meet all of Europe’s heating demands. 

Once a district heating system is established, many heat sources can be connected to it. A great example system is the district heating system network in Aalborg, Denmark. This system is mainly supplied by waste heat from a large cement producer but also recycles waste heat from the local crematorium (Yikes!). The cement production processes require limestone and sand to be burned at high temperatures of around 1500oC. As a result, the Aalborg Portland cement facility has enormous supplies of excess heat. This excess heat has been harvested through heat exchangers etc and transferred into the Aalborg district heating system network. 

In the case of the crematorium, the flue gasses are cooled from around 800 to 120oC and instead of emitting heat to atmosphere, the heat is harvested for use in the district heating system.

It has been quoted that the waste heat from the cement factory and crematorium can heat approximately 24,000 and around 20 to 25 regular Danish homes annually.  

It has been quoted that the waste heat from the cement factory can heat approximately 24,000 homes and the waste heat from the crematorium can heat approximately 20 to 25 regular Danish homes annually.  

As the main function of a crematorium and the cement factory is not to produce heat, the excess waste heat can be seen as a by-product of the respective processes. 

High levels of recycled heat and renewable sources serve to lower the emissions from district heating systems as well as the cost of heat to the end consumer. 

Germany is another good reference of a country utilising district heating systems. In 2018, it was said that approximately 14% of all households were connected to a district heating system. Hamburg stood out at the time as a front runner in district heating. The city had a vast district heating network supplying 19% of all households. This figure was to rise in 2020 by connecting an additional 50,000 households to the district heating network.

We have seen that District Heating Systems can facilitate the use of what could be considered waste heat. In addition to this, DH systems can also facilitate the use of renewable heat sources as well as combine heat and power (CHP) sources. DH technology has been undergoing a transition. Let’s look at a brief summary of the generations of district heating systems. 

First Generation: This generation of district heating system was based on heat distribution by steam. This generation dates from around 1880 to 1930. 

Second Generation: This generation of DH system was based on pressurised hot water with temperatures above 100oC (212 F). This generation dates from around 1930 to 1980.

Third Generation: This generation, like the Second Generation, was also based on pressurised hot water. This generation uses prefabricated, pre-insulated pipes which were buried in the ground and operate with temperatures below 100oC (212 F). This generation dates from around 1980 to 2010.

Fourth Generation: Sometimes referred to as 4th Generation, 4GDH this generation of DH system are designed to operate at lower temperatures (around 65oC (149 F)) and enable a more cost-effective method than that of using fossil fuels. The concept promotes the user of renewable and secondary heat sources such as that from groundwater or heat transfer from waste heat. The lower water distribution temperatures help limit installation costs and reduces ground heat losses. The ability to connect renewable heat sources whilst also harvesting of waste heat helps limit carbon emissions and reduce air pollution. 

The development of DH systems over the years coincides with the advancement of higher insulation standards for buildings. This in turn reduces the heat demand on a DH system. This generation dates from around 2010 onwards. 

Fifth Generation: The evolution of DH systems is to move toward water distribution networks which operate at near-ambient ground temperature conditions. This reduces heat losses to the ground whilst also reduces the overall insulation requirement. This system is then coupled with heat pumps in each connected building. The heat pump extracts heat from the DH system as needed. The same heat pump plant, when operating in reverse-cycle mode, rejects heat back into the DH network when the building requires cooling. This type of operating system permits heat to be shared among buildings from the same DH network. The fifth generation DH system allows efficient heating and cooling in city centres without contributing to air pollution and without emitting CO2

Note, this offers a brief overview on the fifth generation DH system. Further reading is recommended on this topic for a better understanding of the more technical aspects of the system. 

FluidFlow has been used to design and model district heating and cooling systems helping designers to understand operating conditions. Figure 1 gives an overview of a part-plan of a substantial district heating system and Figure 2 an overview of a district cooling system as modelled in FluidFlow. 

District Heating System

 Figure 1: District Heating System (Part-Plan).

District Heating System

Figure 2: District Cooling Plant.

Using a suitable software tool, you can develop a model of the system and evaluate different operating conditions such as fluctuating user demand profiles. FluidFlow can perform such simulations and can be used to model and evaluate other operating scenarios for district heating systems. Take the following as an example case. You may for instance have an existing district heating system which is divided in two grids with each grid supplied from separate powerplant with dedicated pump stations. Each powerplant has two sets of pumps with heat exchangers between them. During the summer period when the flow rate is low (only sanitary water is to be heated) the whole district heating network can be supplied from just one powerplant. It may however be necessary to analyse the complete grid in such operation but with heating included thus with an increased flow. 

Perhaps the system you wish to evaluate is an existing system and you wish to establish potential bottleneck issues through system expansion. A bottleneck can arise when pipes are too small to sufficiently meet end-user needs. This can result in insufficient differential pressure in district heating zones attached to these pipes. Bottlenecks are common in expanding district heating networks as what were previously sufficient pipe diameters become too small as more and more district heating consumers are connected in the network. 

Bottlenecks have in the past been overcome by raising the initial temperature or the initial differential pressure. This however can lead to higher costs and higher heat losses (in the increased temperature scenario). In an attempt to avoid this, replacement of pipes for larger diameter pipes is considered as this increases the flow area reducing the velocity.

There are other techniques available to help resolve bottleneck problems some of which have included adding a heat source to the downstream side of the bottleneck or installation of thermal energy storage. Modelling systems with a suitable software tool can help greatly in identifying system short comings or potential short comings with future system expansion.  A key point here is that the modelling of DH systems can help prevent bottleneck conditions in new or existing systems being expanded. 

If bottlenecks are eliminated, it may be possible to further optimise the system by lowering the supply temperature or reduce the pumping power in the network. This would lead to benefits including lower heat losses and lower pumping power required. 

Another scenario which may result in the occurrence of bottlenecks in DH systems is the transition towards 4th generation of DH systems as these systems are characterised by lower supply and return temperatures than the older more conventional DH temperatures. Lower supply temperatures in DH networks have been proven to exhibit a lot of advantages such as lower heat losses and more straightforward introduction of renewable or waste heat sources. However, lower supply temperature can also introduce some difficulties. The temperature difference in 4th generation systems is around 30oC which is lower than the normal temperature difference of around 40oC in the third-generation systems. Implementing or transitioning from third to fourth generation in an existing system would lead to higher flow rates. This increases the risk of bottlenecks. The same applies to lowering the supply temperature without lowering the corresponding return temperature. 

To avoid too high flow rates in the pipelines, it is therefore important that the cooling among consumers is functioning well and efficiently and the pumping operation is well optimised. 

Research has been completed on the different approaches or techniques which can and have in the past been applied to resolve bottleneck issues. The measures were classified into three groups: 1) Increased Distribution Capacity, 2) Local Supply and 3) Consumer Flow Management. 

Increased Distribution Capacity

Supply Temperature: The approach of increasing the supply temperature can achieve positive results for simplicity and reliability with little or no investment costs. This is however a potentially costly option if it’s not used in small networks and for short time periods.  This approach can result in increased heat losses and reduced efficiency in certain production units which therefore can produce a negative environmental outcome. The maintenance requirement can also increase as a result of higher potential for leaks as well as higher wear on pipes. This of course does depend on how often the supply temperature is increased and for how long. Multiple increases and decreases of the supply temperature can cause more wear on pipes than occasional or less frequent increases and decreases.  

Larger Pipe Area: Although increasing the pipe diameter can improve operating conditions and reliability, it can be complicated by the requirement to excavate and replace existing pipes. This of course can exhibit a considerable investment cost.  

Pumps: The pumps deliver pressure in the DH network and can be used to increase pressure at bottlenecks. Increasing the pump work with an existing pump can produce additional electricity demand with an unfavourable environmental outcome. Different pumping scenarios can be considered in a model, for instance, optimizing or modelling the capacity of the existing pumps, modelling replacement pumps in a system or analysing the effect of having distributed variable speed pumps as opposed to more centralised pump installation. There is evidence to indicate significant electrical energy savings can be achieved by using the latter approach. 

Local Supply

Local heat Supply: This option consists of the addition of a local heat supply into the DH network downstream of the bottleneck. This can be a costly option. This approach does mean less heat and less flow is needed to be transported through the bottleneck location. Consideration also needs to be given to the ongoing future maintenance requirement and cost of connecting a new local supply. 

The addition of a thermal store can also be used to avoid bottleneck problems. Accumulators can be loaded when the DH demand is low and unloaded when demand is high. 

Consumer Flow or Demand-Side Management

At times with large heat demand (mornings and evenings), the space heating in buildings in areas with bottleneck problems may be controlled or kept to a certain threshold value without reducing the quality of service provided to the customers. The intention is to avoid the usage of top load production sources and to reduce energy consumption. 

Note, this discussion simply gives a brief overview of available options to resolve bottleneck conditions. Further reading is recommended on the subject as there is a vast amount of information available on the topic. 

Two-Phase Liquid-Gas Flow Regime

In two-phase liquid-gas flow, the distribution of the respective liquid and gas phases in the pipeline is an important aspect of their description. Their description descriptions exhibit some commonly observed flow structures known as two-phase flow patterns or flow regime. These flow patterns have particular identifying characteristics. Pressure drops and heat transfer coefficients are closely related to the local two-phase flow structure of the fluid. As a result, the prediction of two-phase flow pattern or structure is an important aspect of modelling two-phase pipe flow. 

Some of the flow patterns often encountered in horizontal pipe flow are stratified flow, stratified-wavy flow, bubble flow, plug flow and slug flow, annular flow and mist flow. The following offers an outline description of some flow patterns in horizontal pipelines:

Bubble Flow: The bubble flow regime is characterized by the gas bubbles which are dispersed in the liquid phase often with a high concentration of bubbles in the upper half of the pipe due to their buoyancy. In the case where shear forces are dominant, the gas bubble tend to be distributed more uniformly in the pipe. In horizontal flow, this flow regime typically only occurs at high mass flow rates. 

Stratified Flow: In horizontal flow, this flow regime typically occurs at low liquid and gas velocities and complete separation of the phases occurs in the pipe. The gas-phase travels at the top of the pipe and the liquid at the bottom and the phases are separated by an undisturbed horizontal interface.   

Stratified-Wavy Flow: If the velocity of the gas phase in a stratified condition is increased, waves form on the interface and travel in the same direction as the direction of flow. The amplitude of these waves is dependent on the relative velocity of the two phases. The crests of these waves tend not to reach the top of the pipe. 

Plug Flow: As the gas velocity is increased further, the interfacial waves become larger and can wash to the top of the pipe. The plug flow condition can then occur and exhibits liquid plugs that are separated by elongated gas bubbles. The elongated gas bubbles are smaller than the pipe such that the liquid phase is continuous along the bottom of the pipe below these elongated bubbles. 

Slug Flow: At higher gas velocities, the diameter of the elongated gas bubbles become similar in size to the height of the pipe. The liquid slugs separating the elongated bubbles can also be described as large amplitude waves. 

Annular Flow: As the gas flow rate is increased even further, the liquid forms a continuous annular film around the perimeter of the pipe. This film tends to be thicker at the bottom than at the top due to gravitational forces. The interface between the vapor core and liquid annulus is disturbed by small amplitude waves and droplets can be dispersed in the vapor core. 

Mist Flow: At very high gas velocities all the liquid becomes stripped from the inner surface of the pipe and entrained as small droplets in the vapor phase.

Some of the flow patterns often encountered in vertical pipe flow are bubble flow, dispersed bubble flow, slug flow, churn flow, annular flow, and mist flow. The following offers an outline description of some flow patterns in vertical pipelines:

Bubble Flow: Discrete gas bubbles are dispersed in the continuous liquid phase. These bubbles can vary considerably in size and shape however they are typically almost spherical and are much smaller than the pipe diameter. 

Slug Flow: Increasing the gas void fraction results in the bubbles become very close such that the bubbles collide and combine to form larger bubbles which tend to be similar in dimension to the pipe diameter. These larger bubbles are separated from one another by slugs of liquid. 

Churn Flow: Increasing the velocity of flow results in the structure of flow becoming unstable with the fluid traveling up and down in oscillations but with net flow in an upward direction. This flow regime is an intermediate regime between slug and annular flow. In smaller diameter pipes, churn flow may not develop at all and the flow structure may pass directly from slug to annular flow. Churn flow is a regime which typically should be avoided in two-phase transfer pipelines such as refrigerant piping networks or from a reboiler back to a distillation column because the mass of the slugs may have destructive consequences for the piping system. 

Annular Flow: When the interfacial shear of the high-velocity gas on the liquid film becomes dominant over gravity, the liquid is expelled from the centre of the pipe and flows as a thin film on the wall while the gas flows as a continuous phase through the centre of the pipe. The liquid may be entrained in the gas phase as small droplets to the extent that the fraction of entrained liquid may become similar to that in the film. This flow regime is particularly stable and tends to be the desired flow pattern for two-phase liquid-gas pipe flow.

Wispy Annular Flow: As the flow rate is increased further, the entrained droplets may form transient coherent structures as wisps or clouds of liquid in the central vapor core.

Mist Flow: At high gas flow rates, the annular film becomes thinned by the shear of the gas core on the interface until it becomes unstable and is destroyed such that all the liquid becomes entrained as droplets in the gas phase.  

The flow pattern in a given two-phase pipe is a strong function of the parameters listed below: 

Pipe diameter.

Angle of pipe inclination.

Liquid and gas flow rates.

Liquid and gas physical properties including densities, viscosities and surface tension.

FluidFlow generates a flow pattern map for each pipe in the two-phase liquid-gas flow system. Figure 1 below shows the duty point for a sample pipe which indicates the flow regime is annular mist as noted at the top of the image. This helps the designer identify the flow characteristics in each pipe. 

Flow Regime Fluid Flow


  1. Fluid Flow Handbook, Mc Graw Hill. 
  2. Wolverine III. 

Sulphur fuel limits…..”A Sea Change”.

Sulphur Fuel Limits

It has been quoted that the shipping industry is responsible for emitting around a billion tonnes of carbon dioxide (CO2) per annum. Efforts are being made by the IMO (International Maritime Organisation) to clean up the industry by introducing new limits for sulphur in fuel oil used on board ships. The current sulphur limit in fuel oil used on board ships for main and auxiliary engines and boilers for ships currently operating outside the Emission Control Areas is 3.5 % mass by mass up until 31 December 2019. The new limit set by the IMO will be in place from 1st January 2020 and will be 0.5 % mass by mass. This applies to ships operating outside designated Emission Control Areas. There are some exemptions that can be reviewed in literature published by the IMO. This change in limits will result in significant reductions in the amount of oxide emanating from ships. In summary, ships will be required to reduce sulphur emissions by more than 80 % by 2020. 

Ships can meet the new limits in a number of ways. Low-sulphur compliant fuel oil can be used and are already common in Emission Control Areas. Alternatively, ships can continue to use heavy bunker fuel if scrubbers are installed to capture harmful pollutants from exhaust gas. The latter option would require an upfront investment however, shippers would perhaps have a competitive advantage of being able to use lower cost higher sulphur content fuel. An article published by Hellenic Shipping Hews Worldwide indicates that scrubbers have become a feasible and popular option. In general, there are three different scrubber options available; open loop, closed loop and hybrid systems. Each option has its benefits and drawbacks. In open loop scrubber systems, seawater is used as scrubbing water. Water is taken from the sea, passed through the scrubber and returned to the sea. In closed loop systems, the system is filled with water which is recirculated. The water is therefore supplied from the sea. Only a small amount of water is released to the sea in a closed loop scrubber system. Finally the hybrid option is as you would guess a mix of the open and closed loop systems. This system can operate in both open and closed loop modes.  Figure 1 gives an overview of a closed loop system. 

Ships with scrubbers are considered to be in the minority and as of May 2018, just 494 vessels were noted as having scrubbers installed. This of course is out of circa. 90,000 commercial vessels on the World’s seas.  

Taking a more longer-term approach on the subject, it is considered that LNG (liquefied natural gas) is expected to become a more prominent part of the shipping sector’s fuel supply since it produces almost no SOx or particle matter emissions. Indeed, two recent examples of vessels using LNG are Containerships Polar, a commercial cargo ship and the AIDAnova cruise ship which is the world’s first cruise ship to be driven by LNG. FluidFlow software was used successfully for modelling the piping systems onboard both vessels. The software is used widely in shipbuilding to model fire-fighting, fuel oil, exhaust, effluent, sea water ballast, air flow, open and closed loop cooling systems to name a few.  


  1. International Maritime Organisation.
  2. Hellenic Shipping Hews Worldwide.
  3. Sulphur Fuel Limits
  4. Sulphur Fuel Limits
  5. Sulphur Fuel Limits

Fluid flow and Bayer liquors in the Bayer process

The Bayer Process

The Bayer Process (the process of extracting alumina from bauxite) is an energy-intensive technology, consuming large amounts of fuel and energy. This Alumina or aluminium oxide (Al2O3) is then used to produce Aluminium at a ratio of 2:1. However, did you know that around 75% of aluminium ever produced is still in use in some form today? It is one of the most recyclable materials available and can be recycled over and over, using only 5% of the energy required for producing the original metal, thus somewhat off-setting the high amounts of energy required for its initial production.

The flow of fluids within the alumina process can be simulated with FluidFlow. Friction loss of Newtonian fluids such as water, steam and sodium hydroxide solution are calculated using the normal Darcy equation/Moody diagram method. (The compressible flow calculation is particularly sophisticated taking into account the expansion of the gas and associated temperature and physical property changes. Sonic conditions are also handled. For predicting all physical properties of water FluidFlow uses the IFC-97 formulations developed by the International Association for the Properties of Water and Steam).

Pure Bayer liquors: green/pregnant liquor (or spent liquor) are also Newtonian fluids but with varying physical properties depending on their make-up, for instance, total alkali content etc.  The viscosity, density, vapour pressure and thermal properties would have to be determined separately, but once this is complete, the values can be quickly defined in the FluidFlow fluids database for use in a model. 

These fluids can also be used as the carrier fluids for suspensions in the Bayer process, say when the hydrates are precipitated. These would be regarded as settling slurries and simulated as described above, using an appropriate correlation such as Wilson-Addie-Sellgren-Clift.  

At the very end of the Bayer process we are dealing with a non-Newtonian/non-settling slurry, the “red mud”. Figure 1 shows this type of fluid simulated as a Bingham plastic.  In Figure 1, the pipes are of different diameter.  Consequently, the shear rate is different and therefore the calculated apparent viscosity is different for each pipe. This example also illustrates the concept of yield stress. In the lower pipe, the driving force is insufficient to overcome the yield stress of the fluid and so, there is no flow in the line.

The properties of this red mud were entered into the fluids database where its viscosity correlation was defined as a Bingham plastic.

Figure 1: Red Mud – Bingham Plastic.


Tee Junctions Pressure Loss

Converging & Diverging Flow at Tee Junctions.

Tee junctions are found in most piping systems in many engineering applications. As such, engineers need to determine the pressure losses across these junctions. Steady-state pressure loss coefficients can be calculated using experimental and analytical expressions. However, the calculation of pressure losses across tee junctions can be a tricky task owing to the fact that flow can be merging, dividing, diverging or converging. In some cases, when we have large complex piping systems, we may not even know what flow condition to expect. The calculation or solution of tee junctions can become further complicated when we have reducing tees not to mention changes in fluid phase state. Figure 1 gives an overview of both converging and diverging flow at tee junctions.

Converging & Diverging Flow at Tee Junctions.
Converging & Diverging Flow at Tee Junctions.

This brief blog considers the solution to tee junctions only. FluidFlow also allows engineers to choose the Crane, Idelchik, Miller or SAE pressure loss relationships for bends/elbows. Wye junctions are also accounted for in the software. 

Watch our video on Modeling Tee Junctions

Gas Distribution Systems & Pressure Regulator Stations

Gas Regulator Station.

Natural gas distribution or utility systems are used to deliver natural gas to customers across different sectors including residential, commercial and industrial. Natural gas distribution pipelines tend to operate at higher elevated pressures which enables the gas to travel long distances via a network of underground pipework. This high pressure, although beneficial for delivery over long distances, is unsuitable for end-users and therefore needs to be reduced gas regulator stations to a lower pressure which customers can receive and use.   

Gas pressure regulators are used to maintain a constant and reduced pressure. The regulator adjusts the opening position so that the gas flow through the station is matched to the system demand. As demand on the system is increased, the regulator flow also increases. In the case where demand on the system reduces, the flow through the regulator also decreases as the regulator reacts to the increase in system pressure. In general, the regulator attempts to match the flow of gas through the station to the demand for gas imposed on the system whilst ensuring the regulator maintains the pressure within acceptable limits.

Gas transportation and distribution companies often need to model gas regulator stations (city gate stations, district regulator stations etc) to ensure they are operating within their available capacity. Additionally, it is also helpful to be able to determine if any stations are approaching their full capacity and from this study, identify any upgrade requirements. This analysis can be carried out before the winter months and where upgrades are required, these can then be planned and completed in time to meet the required or anticipated winter load demand.

FluidFlow software includes a regulator component that allows gas transportation or utility companies to calculate the flow of gas and pressure loss through a gas regulator station. Heat transfer can also be considered as well as the Joule-Thomson effect. The software can be used to maintain the accuracy and reliability of hydraulic gas station models created to simulate gas regulating stations and customer station facilities. 

When developing station models, engineers can include the relevant pipes and fittings including filters, gas regulators etc. Bypass lines can also be modeled. 

Gas regulators can be modelled using either the ISA or Universal Equations. 

Figure 1 gives an overview of an example gas regulator station.

Fortis BC in Vancouver have successfully used FluidFlow software for a number of years to model their gas regulator stations. 


  1. Leading Practices to Reduce the Possibility of a Natural Gas Over-Pressurization Event, American Gas Association (AGA).

Gas Distribution Systems Gas Distribution Systems

Fan Performance and Fan Laws

This blog gives a general outline of the rules or laws which can be used to predict fan performance in a given system. Why are the fan laws important? As an example, let’s consider the fan curve typically provided by a manufacturer. This fan curve is usually measured at “standard” or other stated conditions. In real systems, it is unlikely that a fan will be spending its operating life at these identical conditions. Furthermore, suction pressure variations, density changes, composition changes, etc are common and can also affect how the fan will operate in the system.

The fan laws help us estimate how a fan will operate in a system at different speeds, fluid density, impeller diameter, etc. Once we have a basic understanding of these laws, the performance of a fan can be calculated for various conditions.

The performance of geometrically similar fans of different sizes or speeds can be predicted with reasonable accuracy for practical purposes using the fan laws. A higher level of accuracy would require the effects of say, the viscosity of the gas, surface roughness of the fan, scale effect also be considered. However, depending on the level of accuracy required, for many fan calculations, this may not be necessary.

One point to note is that the laws apply to the same point of operation on the fan curve. They cannot be used to predict other points on the fan’s curve. The fundamental fan laws governing fan performance are generally only valid for a fixed system with no changes in airflow characteristics in the system or changes in aerodynamics. The term “system” refers to the combination of ducting, filters, grilles, dampers louvres, hoods, etc through which air is distributed.

As we know, the movement of air through a system causes friction/resistance between the air molecules and their surroundings and any other air molecules. Energy is therefore required to overcome this resistance. The faster the air moves through the system the greater the resistance imposed to flow and the more energy required to deliver the air through the system. This energy is described in terms of pressure. In general, the pressure required to overcome the resistance is referred to as static pressure. The pressure that results in the air/gas velocity is described as velocity pressure and the combination of these two values is often referred to as total pressure.

Fans or blowers are often installed in the ventilation or industrial process systems to overcome the resistance. Fan performance is often represented in the form of fan curves. The curves are based on a specific set of conditions which typically include speed, volume, efficiency, static pressure and power required to drive the fan at the given set of conditions. Figure 1 provides a typical illustration of fan curves. The point where the static pressure curve intersects the system resistance curve represents the duty-point for the fan.

Fan Performance Curve
Figure 1: Manufacturer’s Fan Performance Curve.

As noted earlier, as the airflow is increased in any fan system, the system resistance also increases. In a fixed system, it is said that the pressure required/system resistance varies with the square of the volume of air flowing through the system. The system resistance curve can be developed by determining the pressure required over a range of system flow rates. This resistance curve can then be plotted on the fan performance curve (also known as the fan capacity curve) to identify the actual duty-point. This is shown as point “1” in Figure 2 where the fan curve N1 and system resistance curve SC1 intersect. This duty-point is at airflow Q1 which is delivered against pressure P1.

System Resistance Curves
Figure 2: System Resistance Curves.

Fans operate along a performance curve as provided by a manufacturer for a given fan speed. If we wish to reduce the air flow in the system, we could for instance partially close off a damper in the system or reduce the fan speed. Partially closing a damper will result in a new system resistance curve. This is shown as system resistance curve SC2 where the required pressure increases for any given air flow. The fan will now operate at duty-point 2 to provide the reduced air flow Q2 against the higher pressure P2.

On the other hand, we can reduce the fan speed from say N1 to N2 to reduce the airflow in the system and keep the damper in the fully open position. Under these conditions, the fan will now operate at duty-point 3 to provide the same airflow rate Q2 but at a lower pressure.

Therefore, reducing the fan speed is a much more energy-efficient approach to reduce airflow since less power will be required resulting in less energy consumption.

Fan Laws

In general, the fan laws are typically used to calculate changes in flow rate, pressure and power of a fan when the size, speed or gas density is changed. In the fan laws outlined in Table 1 below, the subscript 1 represents the initial existing condition and subscript 2 represents the desired calculated condition.

The illustration in Figure 3 helps illustrate the effects of changes in fan speed.
Pressure & Power due to changes in fan speed
Figure 3: Effects on Flow, Pressure & Power due to changes in fan speed.

The fan laws are a group of equations used to determine the effects of changes in the fan operating speed, the fan diameter or the density of the air in the system. The performance of a centrifugal fan, axial fan or blower is often given as a series of pressure, efficiency and shaft power characteristic curves plotted against air flow rate for specified values of speed, air density, and fan dimensions. It is, therefore, useful to determine the operating characteristic of the fan at other speeds and air densities. Using the fan law relationships, families of fan curves can be developed for operating the fan at different speeds, etc.

The fan laws can also be utilized to consider test results obtained from smaller prototype fans to predict the performance of larger fans which are of course geometrically similar.

Knowing the performance of a given fan under set specified operating conditions, variations in the performance can be predicted according to the fan laws. It should be noted however that adding or removing components of a fixed system such as dampers or incurring density changes will create an entirely different system resistance curve. Also, it’s worth noting that changing fan accessories such as inlet dampers, inlet boxes will change the fan performance curve from the standard. This should, therefore, be considered before considering or applying the fan laws.

As part of the system design, the fan laws can be quite useful in determining alternate performance criteria or in establishing a minimum and maximum range. In the event that “safety factors” have been applied to system design calculations, it is worth noting that, based on the fan laws, a 10% increase in volume will result in a 33% increase in power requirement. Due consideration is therefore recommended on evaluating any applied “safety factors” against the actual cost penalty incurred.

In general, using these rules or fan laws, once we know the performance of a given fan under set specified operating conditions, variations in the performance can be predicted with reasonable levels of accuracy. FluidFlow software considers the effects of compressibility, suction pressure variation, etc whilst also solving the fan laws ensuring a high level of accuracy.


  1. Bureau of Energy Efficiency.

Pump Sizing – The Effects of Oversizing Centrifugal Pumps

Pump Sizing: We’ve probably all come across the topic oversizing of centrifugal pumps in the past but why can it happen and what are the effects? This blog takes a look at the condition of oversizing centrifugal pumps. 

Pumps can be oversized for a number of reasons. Sometimes a pump is sized early-on in the planning or concept-design stage using estimated values for pipework and fittings losses. A degree of error can occur at this stage as it would be necessary to guesstimate the pipe lengths, quantity of fittings etc as well as the possibility of published literature having different reference values for flow resistance coefficient or fittings losses. As a consequence, the dynamic pressure losses can be questionable. On the other hand, the static pressure differences are often usually better defined. 

Another potential reason for pump oversizing is perhaps the system being designed will potentially expand in a few years and the pump size calculated won’t meet this future additional demand. A “safety margin” is sometimes added which can often be an arbitrary value. A larger pump is then procured and installed with the view to serving the future extended system. 

Other reasons may include, the selected pump was the only pump in stock by the supplier and a pump was required urgently – deal was agreed with the vendor and the pump procured, a pump was perhaps selected from the spares inventory due to budget constraints or, a new replacement pump was selected for an existing system. This new pump matched the pump already installed however, the existing pump was already over-sized. Another reason might be the designer over-sized the pump based on expected build-up of corrosion products on the pipe interior and as such, would over time increase the pump total head requirement. One final point to consider is, as a pump wears, there can be a reduction in head and capacity. This may present the argument of initial additional “safety factor” to increase the capacity and head with the view to counteracting the eventual effects of wear. 

Some of the above reasons may seem reasonable however, it is sometimes the scale of “safety factor” applied which can lead to knock-on negative effects on the final pump operating conditions, capital and running costs etc. This can be compounded by a vendor selecting a pump which has another “safety factor” to be certain there there is “enough pump capacity to do the job”. 

Let’s take an example scenario to illustrate the effect of pump oversizing. 

A designer has established a design flow rate of 911 m3/h at 26.0 m fluid. Believing that using a safety factor will ensure we reach duty-point D as shown in Figure 1, the designer erroneously adds 15.5 m fluid head to obtain a total head of 41.5 m fluid. Figure 1 gives an overview of the performance curves for the system under consideration.

Centrifugal Pump Performance Curves
Figure 1: Centrifugal Pump Performance Curves.

Assuming the designer actually needs a pump to operate at 911 m3/h and 41.5 m fluid head, a pump vendor selects a pump with a curve A, B & C. The pump curve intersects the system curve at the best efficiency point. However, the actual system curve is E, D & C and the pump will run at duty-point C rather than B. This is because the pump seeks equilibrium with the system and operates at the intersection of the pump curve and system head curve. At point C, the pump will produce a flow rate of 1120 m3/h. So, not only will the designer achieve different operating conditions, the pump is also operating at a lower efficiency point on the pump curve and therefore, exhibits a higher operating cost.   

In order to achieve the original intended design flow rate of 911 m3/h, the system valve is gradually closed/throttled. This will “steepen” the system head curve since we are adding resistance to the system. The pump then produces 911 m3/h at 41.5 m fluid. However, the head requirement at 911 m3/h is 26.0 m fluid. The pump therefore actually produces 41.5 m fluid and 911 m3/h but delivers only 26.0 m fluid 911 m3/h to the system. The additional 15.5 m fluid head is wasted across the valve as heat and noise. 

What are the cost implications of oversizing a pump? 

Now, let’s make a quick “back-of-an-envelope” calculation for this example and assume that this pump is 60% efficient at duty point C. 

911 * (41.5 / 360) =  105.0 kW produced and 105.0 / 0.6 =  175 kW required to do this.

Now, let’s consider the duty point at point B and consider this pump is 68% efficient at this duty point. 

911 * (26.0 / 360) =  65.8 kW produced and 65.8 / 0.68 = 96.7 kW required to do this.

The difference is 175 – 96.7 = 78.3 kW which is being used to pump against the throttled discharge valve. Note, a discharge throttling valve is often used to overcome the effect of pump oversizing since it’s a simple solution to complete on site however, it comes at a cost of wasteful energy consumption. If this were a pump which was running 24 hours a day, this would mean we would have 24 * 365 * 78.3 = 685,908 kW per annum of wasteful energy use. How much are you spending on a kilowatt hour of electricity? Let’s assume it is $0.05 c/kWh. That simply means we are wastefully spending 0.05 * 685,908 = $34,295 per annum on operating costs for this pump alone. 

We can see from Figure 1 that the pump duty point would actually be closer to the end of the pump curve than intended but what are the practical implications of operating a pump towards the end of the curve? Firstly, pumps are designed to operate near their highest or best efficiency point. When they operate towards the right hand side of the curve, the likelihood of cavitation increase. The suction and discharge pressure can fluctuate wildly pushing the impeller about because of the out-of-balance forces. The pump shaft rattles, packing becomes worn, mechanical seals become damaged, bearings are destroyed by hammering (brinelling) and shafts are bent. In operation, pumps will be noisy, vibrate and shake, get hot and the microjets of fluid ejected by imploding bubbles will erode the impeller. As a result of these conditions, the pump will unsurprisingly tend to require a lot of maintenance and will fail regularly. 

How can I tell if my existing pump is oversized?

There are a number of factors which can suggest that an existing pump is oversized for the installed application. Let’s look at a few: 

Highly Throttled Flow Control Valves

Control valves installed in system featuring centrifugal pumps tend to stay in restrictive valve positions which in turn forces the pump to work against a high back pressure. Since this is higher than the pressure required for the pump to achieve its best efficiency point (see previous blog-post titled “Operating Efficiency of Centrifugal Pumps”), the pump runs inefficiently and at risk of increased bearing wear. 

Excessive Noise

An oversized pump can cause excessive noise and vibration of pipes which over time can loosen flanged connections and joints as well as cause fatigue of pipe welds. This can lead to premature wear and failure. Surprisingly, excessive noise can sometimes go unnoticed since the site operators may simply be used to a given noise-level on site and as such, it is considered since the plant is operational and in sound working order.  

Regular Replacement of Bearings and Seals

Oversized pumps tend to generate high backpressures which results in accelerated bearing and seal wear. If a pump requires frequent replacement of bearings and seals, this could be an indication that the pump is oversized. 

Heavy use of bypass lines

Bypass lines are sometimes used to manage excess flow and allow a pump to operate closer to its best efficiency point (BEP). Whilst this may improve the reliability of the pump it also results in wasteful energy consumption since the energy required to move the fluid through the bypass lines is wasted. If a system has a large number of bypass lines, it could be in indication of that the system has scope for improving its energy efficiency. 

Intermittent operation

Pumps are sometimes used to maintain a steady liquid level in a tank, draining and filling as required. This constant cycling of powering up and down reduces the lifespan of the pump motor controller and pump assembly which means the pump will be more at susceptible of intermittent operation. Oversized pumps tend to generate higher friction losses since they force a greater flow rate of fluid through the system that it needs to. When these problems get progressively worse, the pump will begin to operate sporadically before failing completely. 

What can I do ?

Potential steps to overcome or resolve oversized pumps include: reduce pump operating speed, reduce the pump impeller diameter or replace pump with correctly sized and selected pump model. This would represent some of the more energy efficient means of resolving the issue. 

Another option would be trimming the pump impeller as this reduces energy consumption due to excessive head from an oversized pump. An impeller can be trimmed with the view to getting the control valve to operate in a more desirable position, i.e. between roughly 40 to 60 % open. If for example the control valve has an initial operating position of say 25%, then the impeller can be trimmed to obtain a more desirable position such as 55% open. The pressure drop can be estimated using the control valve equation by getting the Cv for each valve position. It should be noted however that this does little for over capacity of the pump sizing. Trimming the impeller can in some cases lead to lower pump efficiency. 

Perhaps the pump can be provided with a variable frequency drive (VFD). This can be a good solution to removing the control valve and improving efficiency. In general. VFD’s are only really suited to high friction head systems and not high static head systems. A useful rule of thumb is that, VFD’s are typically only justified when the dynamic head represents 50% of the total head in the system. In  a system which serves many differential pressure control valve systems, VFD’s may not be a feasible option. 

In summary, the effects of pump oversizing include: operating at excess capacity can require a higher NPSHr, cavitation leading to reduced efficiency and premature component failure, higher power consumption, higher ongoing running costs, higher capital purchase cost, internal loading and hydraulic radial thrust, vibration and dehydration. 

Pump Sizing Pump Sizing Pump Sizing Pump Sizing Pump Sizing

To View our Blog on “Operating Efficiency of Centrifugal Pumps”


UC Davis OneLoop Pod Hyperloop competition

Good luck to UC Davis OneLoop Pod in the SpaceX 2019 Hyperloop Pod Competition on July 21.  Proud that FluidFlow was crucial in designing a potential prototype of the transportation of the future!

The Hyperloop is a novel form of low-cost, fast transportation, consisting of levitating pods propelled through a vacuum tube at transonic speeds. The concept is projected to be able to deliver passengers from San Francisco to Los Angeles in 30 minutes for just $20.  To further research and innovation at the university level, Hyperloop’s ideator Elon Musk started an international competition for student teams through SpaceX . UC Davis is one of the top 21 teams that will compete in July 2019.

The OneLoop Pod is an innovative, high-tech vehicle that will represent UC Davis in the 2019 Hyperloop Pod Competition run by SpaceX. Hyperloop technology aims to revolutionize transportation by combining high-speed travel with vacuum tube technology. This year’s OneLoop pod features a high-performance programmable logic controller, a lightweight 6061 Aluminum chassis, mutually redundant pneumatic friction brakes, independent quarter-car suspension modules, a three-axis lateral stability mechanism, and a Nitrogen cold-gas rocket thruster.

The primary propulsion system consists of four composite, high-pressure tanks filled with Nitrogen gas, a series of redundant air-operated valves, and a converging-diverging aluminum nozzle.

On top of the primary system, a high pressure regulator, additional high-pressure valves, a fill port, and a flow restrictor comprise stations for both filling and scheduled relief of the gas in the tanks.

Find out more at the links below:

Two-Phase Safety Pressure Relief Valve Sizing.

Relief Valve Sizing Model

The mass flow of fluid discharged through a safety pressure relief valve is typically a liquid or vapor however, it can be a combination of both liquid and vapor or in other words, two-phase flow. When considering pressure relief applications, it can be common for all or part of the liquid portion of the fluid to flash or change to its vapor phase-state as the pressure drops. The ratio of liquid to gas in the flowing fluid can be an important factor in determining the required relief valve orifice area. 

Since two-phase flow generally has a decreased flow capacity compared to vapor-phase or subcooled (non-flashing) liquid flow, a greater relief orifice area is generally required for two-phase flow. Indeed, literature sources occasionally quote that the required relief area for two-phase flow is around two to ten times the area for single-phase flow. The larger area requirement results from the fact that when we have two-phase flow conditions, the liquid partially blocks the relieving area flow path of the vapor by which most of the energy is removed by evaporation from a vessel. 

Oversizing of pressure relief valves can be as detrimental as undersizing. Oversizing a safety relief valve with two-phase flow can have dangerous consequences. Excessive fluid flashing can occur on the downstream side of an oversized relief valve which can cause the back-pressure build-up to the point that the relief device function becomes impaired. The result could potentially be a catastrophic vessel failure. 

The selection of sizing method to be used for a two-phase flow application is a decision made by the designer. It is therefore vitally important that the designer has detailed knowledge of the process conditions. The use of an appropriate software tool can also help greatly with the relief valve orifice sizing task.

The design of two-phase safety relief valves can be a complex undertaking which is why an experienced expert should be consulted and a suitable software tool can help enormously with this challenge. 

There are a number of published codes of practice and standards which discuss the topic of relief orifice sizing for two-phase flow conditions. These documents may help the designer choose the appropriate sizing methodology. Further reading of API 520 Part I and EN ISO 4126 will provide some useful information regarding suitable equations for a given design condition. 

API 520 Standard Part I details an example hand calculation which considers a sizing case for propane with a volumetric flow rate of 100 gal/min and relief valve set pressure of 260 psig. The fluid temperature at the relief valve inlet is 60 F, the downstream back pressure is 10 psig and permitted accumulation is 10%. Based on the design conditions described in the document, the orifice area has been calculated at 0.208 in2. From this calculation, the standard orifice size “F” was selected (0.307 in2). 

This example was recreated in FluidFlow software (Figure 1) and in doing so, a number of assumptions are required since the standalone hand calculation in API 520 doesn’t take into account the effect of connected pipes. Pipe sizes, lengths etc are therefore not described in the document. The model utilises an upstream and downstream pipe connected to the PRV and the design conditions are defined for the system. 

Relief Valve Sizing Model

Figure 1: Relief Valve Sizing Model. 

We have seen that the API Standard arrived at a calculated orifice size of 0.208 in2 whereas FluidFlow has calculated a size of 0.207 in2 (Figure 2). When using the API Pressure Loss Model in FluidFlow, the software automatically suggests the next closest standard size match which in this case, is orifice size F matching that of API 520 Part I.

Relief Valve Sizing Results
Figure 2: Relief Valve Sizing Results. 

The example hand calculation makes reference to a density of 31.92 lb/ft3 for the propane at the inlet of the relief valve. Figure 2 notes that FluidFlow has calculated a density at this point of 31.64 lb/ft3

The software has also detected flashing has occurred across the PRV and applied the appropriate pressure loss correlations. The fluid phase state is clearly visible at both the inlet and outlet of the relief valve. 

In some cases, industry standard documents recommend that the relief valve manufacturer is contacted directly for advice for sizing relief valves for two-phase flow conditions. 

Ensuring an emergency safety relief system design can either avoid or accommodate two-phase vapor-liquid flow is a particularly important aspect of the design process. 

The use of closed-loop piping systems in pharmaceutical processes

Most pharmaceutical organisations utilise process cooling plant as part of the manufacturing processes. The cooling water is used for tablet formation, batch processing, R&D, liquid sterilisation, ointment and cream cooling etc. In recent years, many pharmaceutical companies have opted to use closed-loop (the subject of an earlier blog post) cooling towers as opposed to open-loop systems. There are a number of reasons for this as in cooling tower systems, the continuous operation of fans and pumps makes these systems quite energy-intensive resulting in high operating costs. The systems can also suffer from solid deposits, gases, scale accumulation on heat exchangers, oxidation, microbiological growth, algae, bacteria/legionella, etc. it is, therefore, necessary to enforce suitable and costly and intensive maintenance procedures and chemical treatment to prevent these undesirable conditions arising.

Traditional cooling towers utilise evaporation to reduce the process water temperature. In the case of closed-loop sytems, the process water is cooled using ambient air. The latter system produces significant energy savings and as closed-loop systems reuse the process water, considerable reductions in water consumption is achieved. This helps pharmaceutical organisations to get closer to the required or target sustainability goals. It has been noted in published literature that, when using closed-loop process water systems in lieu of open-loop systems, companies can save up to 98% of process cooling water due to the recirculation of water around the loop. 

The reuse and recirculation of the clean water also eliminates notable costs associated with consumption disposal and treatment. Further efficiencies can be achieved from the use of closed-loop processes cooling systems as in many climates, chillers can be turned off permitting free-cooling during winter months. The control system can monitor outdoor air temperature and shut down the refrigeration compressors when appropriate. These compressors are large energy consumers and as such, having the ability to shut down these units when the climate permits, saves considerable energy thus reducing operating costs. Energy reduction of up to 95% through effective implementation of such a control strategy has been quoted in published literature. 

The use of variable speed fans and pumps further increase the savings through the optimised use of energy. 

Many pharmaceutical sites also utilise pumped purified water loops. Whilst more modern facilities are designed to include energy efficient systems and energy saving measures, older systems can also offer energy saving opportunities. The following example from a pharmaceutical site is of an existing purified water loop system which contains a single circulation pump which provided a maximum flow rate of 56 m3/h at 8 bar pressure. The general operating demand was 40 m3/h at 8 bar pressure. The total length of the piping is approximately 246 M with a rise of approximately 25.0 M on the return leg to the tank.

It was considered by the client that the existing pump was not matched to the system and coupled with the extensive daily system operating hours, was contributing to the high system electrical costs experienced on site. 

A baseline model was developed for the system and solved at a design flow rate of both 40 and 56 m3/h. Figure 1 provides an overview of the model.

Figure 1: Purified Water Loop (Design Flow Rate of 56 m3/h).

Based on a design flow rate of 40 and 56 m3/h, the calculated duty pressure rise is 31.56 and 66.29 m fluid respectively. 

The existing circulating pump was then assigned to the model and it was seen that the pump generated a flow rate of 61.9 m3/h with a duty pressure rise of 78.67 m fluid. This performance was based on the pump speed of 2900 rpm and an impeller diameter of 255mm. The duty power requirement under this operation was 24kW. Figure 2 provides an overview of this simulation. Reducing this pump speed to 2656 RPM reduced the flow rate to 56 m3/h. The associated power requirement was 18.4 kW.  

Figure 2: Purified Water Loop – Original Pump (61.9 m3/h).

A proposed alternative pump as put forward by a pump vendor was then modeled in the system (Figure 3). This pump produced a flow rate of 56 m3/h. The power required for this pump was 15.2kW. Reducing the speed of this pump to achieve a flow rate of 40 m3/h lowered the power requirement to 5.8 kW, much lower than the original pump installed in the system.

Figure 3: Purified Water Loop – Alternative Vendor Selected Pump (40 m3/h).

 This example from a real pharmaceutical site demonstrates how energy savings can be achieved in existing purified water systems in pharmaceutical sites. Cooling water distribution as well as other utility systems can be modeled thereby identifying energy saving opportunities.  


  1. Pharmaceutical Manufacturing. 

Two Phase Flow Liquid-Gas System Design

The design of two-phase liquid-gas pipe flow systems is a complex process. It can be common practice for engineers to use template design calculation sheets for two-phase systems. These calculation sheets often utilize rules-of-thumb as part of the template for design. This can result in the failure to acknowledge inherent differences from one project to the next.  In large-scale projects such as geothermal systems, errors in system design such as underestimating friction losses can become magnified resulting in poor design with inherent operational difficulties.

In this paper we will make a hand calculation for a pipe system that has both vertical and horizontal pipe runs and then compare the results to those that can be obtained from a software based solution. 

This piping system has 358ft of level pipe, three vertical rises of 10ft each and one vertical rise of 50ft. We wish to evaluate the type of flow and expected pressure drop. 

Table 1: Fluid Physical Property Data.

Pipe Data: 3 Inch, Schedule 40 Stainless Steel (I.D. 3.068 in). Pipe Relative Roughness: 0.000587. 

A hand calculation has been completed for this system, details of which are outlined below. 

This calculation is based on the gas having a viscosity of 0.00127 cP and a density of 1.23 kg/m3.

To make a comparison, a model has been created in FluidFlow software. For convenience, based on the fluid physical properties, the liquid and gas used in the model is water and air. It is worth noting that the viscosity of air and water in FluidFlow is 0.018 and 1.13 cP respectively. This represents a very slight difference to the hand calculation.

FluidFlow Solution

The first step is to build the model using the above data. The model should appear as per Figure 1 below. Note, two Known Flow inlet boundaries have been defined, one with air and one with water. The connecting pipe has been set to 0.5m in length, 3 inch in diameter and the statues of these pipes (nods -1 & -3) have been set to “Ignore Pressure Loss” as, these pipes are only provided for model connectivity purposes and do not form part of the overall calculated system.

Figure 1: System Model.

The system consists of a total length of 358ft of level pipe. As we have a “stepped” system i.e., changes in elevation, we have simply split the 358ft into 5 equal level pipe segments of 71.6ft each. Details of the various pipe lengths are set out in Table 2. 

Table 2: Pipe Data.

The node elevations have also been set in accordance with the system design criteria. Details of which can be seen in Table 3 below. 

Table 3: Node Elevation Data.

This system has been calculated using the various different two-phase pressure loss correlations available in FluidFlow. Table 4 provides details of the calculated results which are compared to the results provided in the literature. 

Table 4: Table of Calculated Results.

Comparison of Results 

  1. The first option selected is Whalley Criteria. This option lets the software select the appropriate method from the three correlations available when using this approach i.e., Friedel, Chisholm Baroczy & Lockhart Martinelli. In this example, FluidFlow has analysed the fluid physical properties (fluid viscosity ratio and mass flux) and has applied the Friedel correlation. Note, this correlation uses different equations for horizontal and vertical pipes. We can see that the results compare well with that in the hand calculation. 
  2. Chisholm Baroczy: This approach dates back to the 1970’s and uses a two-phase multiplier. This is one of the more simplistic two-phase correlations. However, the results compare well when using this approach. 
  3. Lockhart Martinelli: This forms the basis of the solution in the hand calculation. In general, it is accepted that this model should only be applied to horizontal pipe flow. This is also one of the more simplistic two-phase correlations and also dates back to the 1970’s. Again, the results compare well when using this approach. 
  4. Drift Flux: This approach is recommended for modelling vertical or inclined pipelines. 
  5. Beggs Brill: Recommended not to use this method for vertical upward flow as it under predicts pressure loss. It is difficult to assess the full effect of this in this particular example as only a fifth of this system consists of vertical pipework, i.e. most of the pressure loss occurs across the horizontal pipe segments. 
  6. MSH: This approach is more useful when considering heat transfer and modelling single component fluids such as refrigerants. This method tends to lose accuracy when high vapor quality conditions are experienced in a system. This particular system has an outlet vapor quality of 0.75 (75%) and as such, it is considered that the results obtained using MSH may have a slightly lower level of accuracy when compared to, for example, the Friedel method. 
  7. HEM: This method uses a simplified approach and averages the fluid density values. It also assumes the velocity, pressure and temperature between phases are equal. Due to the simplification of the modelling approach, it is recommended that this option should only be selected as a means of checking or validating results from the correlations noted above. It is also clear from the HEM approach that the total system pressure drop is significantly lower than that provided in the hand calculation and also when using other available loss correlations in FluidFlow. 

Note, interestingly, the three approaches of Friedel, Beggs Brill and MSH which take into account flow orientation or flow regime offer a solution which is closest to the hand calculation example. 

FluidFlow generates a flow pattern map for each pipe in the system. Figure 2 below shows the duty point for a sample pipe in this system which indicates the flow regime is annular mist which correlates well with the hand calculation. 


The literature calculation is based on the gas having a viscosity of 0.00127 cP and assumes gas ideality. For simplicity, the model has been developed using air and water which has a viscosity of approx. 0.018 and 1.13 cP respectively (based on 15oC). This will therefore have a slight effect on the calculated results. 

The “hand” calculation is based on ideal gas conditions. FluidFlow does not assume gas ideality but solves for real gas conditions using an equation of state and hence, provides more accurate result. 

Based on the above, it is considered that the results provided by FluidFlow correlate well with the hand calculation and offers an accurate reflection of the system operating conditions. It is also considered that the Friedel correlation may be best suited for this particular application owing to the combination of both vertical and horizontal pipes. 

Note, solving this system in a suitable software package allows the designer to consider different correlation approaches and also identify the flow pattern map for each pipe in the system. 

The design of two-phase piping systems may pose a daunting task to the designer. However, having a useful and easy to use software tool can help eliminate the likelihood of data entry errors, enable better understanding of the operating performance the system, help identify unstable operating conditions and also, consider different design conditions without the need to refactor any hand calculations. Furthermore, once a design has been finalized, you can revisit in the future should the operating system require expansion or alterations. 


“MAN Diesel & Turbo’s activities in the power plant sector are based on a well- established range of diesel engines and a rapidly growing gas engine offering. The products range from small emergency power generators to turnkey power plants with outputs of up to 400 MW. We use the liquid, gas, and two-phase modules of FluidFlow and make extensive use of its simulation capabilities for engineering subsystems as fuel gas lines or cooling water pressure systems, including for the development of new systems. Before we bought we carried out extensive product research and chose FluidFlow because of its completeness and value for money. We have had to make
occasional technical support calls and have been impressed by the responsiveness. We have found Flite Software to be extremely knowledgeable and helpful, really excellent.”

Norman Kurth, System Engineering, MAN Diesel & Turbo, Germany

Geothermal Power Plants.

Geothermal power plants utilise hot water and steam from deep underground dry steam or hot water wells which is piped upwards and used to generate electricity. Technologies used in geothermal power plants typically includes dry steam, flash steam or binary cycle power plants. 

Dry Steam Plant

In dry steam plants, hot geothermal steam is piped from geothermal reservoirs which then drives turbines. The turbine powers a generator which produces electricity and adds to the power distribution network. In dry steam plant, the steam is transferred from the well to the powerhouse via the gathering piping system. This piping system generally consists of one or more large diameter mains pipelines which collects the steam from the smaller diameter pipelines connected to each wellhead. As the main pipe approaches the powerhouse, the mass flow rate increases as each wellhead is connected into this common mains pipeline. One of the main considerations in the design of the gathering system is the selected diameter of the pipelines. For steam transportation pipelines, an allowable maximum velocity of circa 50 to 60 m/s for large diameter lines (>300mm pipes) whereas for smaller diameter pipes the velocity of circa 20 to 25 m/s is recommended. These velocity limits are recommended as the presence of water droplets or fine solid particles carried by high velocity steam can cause erosion of valve seats, bends as well as any other exposed parts. 

At the well, there is a valve arrangement with steam purifier (inline axial centrifugal separator) which is designed to remove water droplets and particulate matter from the steam before it enters the piping system. The steam pipelines are fully thermally insulated and typically include expansion loops to accommodate any pipe movement. Steam traps are positioned strategically along the pipeline to remove condensate which is then transferred by a separate piping system to holding ponds before being reinjected. As the main gathering line approaches the powerhouse, a safety relief valve station is provided to permit the safe removal of steam in the event of turbine trip. Safety pressure relief valves can be automatically sized in FluidFlow to ISO & API Standards. Experience shows that it is better to operate the wells in a steady open state instead of changing between open and closed positions. 

Flash Steam Plant

Single-flash steam plant is said to be the most widespread type of geothermal power plant. This plant is often the first power plant type installed at a newly developed liquid-dominant geothermal field. In geothermal wells which produce a two-phase mixture of liquid and steam, the single-flash plant offers a simple method of converting the geothermal energy into electricity. The mixture is firstly separated into two distinct liquid and steam phases with a minimum loss of pressure. This step is completed in cylindrical cyclone separators where the water and steam phases separate due to their inherent density differences. Separators are often positioned at the powerhouse, at satellite stations in the geothermal field or at the wellheads. 

One of the main considerations in the design of the gathering system is the pressure loss in the steam lines from the wellhead to the powerhouse. The pressure loss in steam pipelines is a function of the configuration of the piping system such as pipe length, diameter as well as the mass flow rate and physical properties of the steam. The most critical design parameter is the pipe diameter since the pressure drop is the function of internal diameter to the fifth power. Installing larger diameter pipes reduces the pressure loss however the extra cost of installing larger diameter pipelines may not be feasible. An optimum pipe diameter should be selected based on a thermodynamic study whilst also giving due consideration to capital cost. Software tools can assist with selecting optimum pipe diameters and FluidFlow offers three automatic pipe sizing approaches; Economic Velocity Sizing, Design Velocity Sizing and finally, Design Pressure Gradient Sizing. 

The pressure loss in liquid pipelines is less of a concern since the liquid shall be disposed of by injection. However, high pressure losses may require pumps to maintain sufficient reinjection pressure. 

The pressure loss in two-phase (steam-liquid) pipes is a much more complex phenomenon and difficult to predict accurately using tools such as spreadsheets as this approach doesn’t accurately consider the interaction of connected equipment. The solution of two-phase pipe flow systems is further complicated by the presence of different flow patterns which can change with pipe orientations etc. The mechanism for pressure loss is different for each flow pattern and therefore the use of a suitable software tool can assist greatly in this regard. FluidFlow includes several two-phase correlations and determines the flow pattern in each pipeline and solves accordingly. The software also plots the flow pattern map for each pipe segment. This allows the designer to identify the flowing conditions, flow regime and also identify how close the pipeline is to undesirable flow regimes such as slug flow conditions. 

The double-flash steam plant is an evolution of the single-flash plant in that it typically can produce around 15 to 25 % additional power output for the same geothermal fluid conditions. The main difference is the presence of a second flash process which takes place on the separated liquid leaving the separator. This process generates additional steam at a lower pressure than that of the primary steam. This system therefore extracts even more energy by comparison to a single-flash system.     

In double-flash steam plant, several different piping arrangements (in comparison to single-flash plant) can be utilised to provide the additional flash process. They can generally be summarized as follows:

  • Two-phase pipe connections from each well to satellite separation stations in the field. High and low-pressure steam lines are then derived from each satellite separation station and piped to the powerhouse. Hot water pipelines from the satellite stations is then reinjected to the wells. 
  • Two-phase pipe connections from each well to satellite separation stations in the field. High pressure steam lines and hot water lines derived from each satellite separation station and piped to the powerhouse. The power house includes separators where short low-pressure steam pipelines are connected to the turbines and hot water pipelines from the powerhouse are piped to the injection wells. 
  • Separators at each well head with high and low-pressure steam lines piped to the powerhouse. Hot water is piped from each production well to the injection wells. 
  • Two-phase pipe connections from each well piped to the powerhouse with separators located at the powerhouse. This system features short lengths of high and low-pressure steam lines to the turbines. The hot water pipelines are piped from the powerhouse to the injection wells. 

A combination of the above arrangements may also be considered depending on the field conditions. The final selected design solution will take account of geofluid conditions, field conditions, well positions, topography position of powerhouse etc. 

Figure 1 provides an illustration of a double-flash two-phase flow system as modeled in FluidFlow. 

Figure 1: Geothermal Double-Flash Two-Phase Flow System.

Flash steam plants are very common and use fluid temperatures in the range of circa. 182oC. 

Binary Cycle Plant

Binary cycle power plants use the warm geothermal fluid (typically below 204oC) to heat a secondary fluid (such as isobutane) which exhibits a much lower boiling point. The fluids are kept separate with heat energy transferred from the geothermal fluid to the secondary fluid via a heat exchanger. This causes the secondary fluid to flash into steam which drives the generator turbines.  

Binary cycle power plants are essentially closed-loop systems and virtually only water vapor is emitted to the atmosphere. The most common geothermal resources exhibit temperature below 149oC and as such, a considerable proportion of future geothermal electricity could come from binary cycle power plants.

A combination of both flash steam and binary cycle technologies has also been utilized effectively to take advantage of both technologies. In this plant, the flashed steam is first converted to electricity with a backpressure steam turbine and the low-pressure steam exiting the turbine is then condensed in a binary system. This allows for the effective use of air cooling towers with flash applications and takes advantage of the binary process. The flash/binary system has a higher efficiency where the well field produces high pressure steam while the elimination of vacuum pumping of non-condensable gases allows for 100 percent injection. Examples of these systems include the Puna Geo Venture facility in Hawaii where the geothermal fluid is over 316oC, the Upper Mahiao geothermal facility in the Philippines as well as several power projects in New Zealand.

A cooling system is essential for the operation of modern geothermal power plants. Cooling towers prevent turbines from overheating and prolong facility lifespan. Many geothermal power plants use water cooling systems. Figure 2 provides an illustration of a cooling water system as modeled in FluidFlow. 

Figure 2: Cooling Water System.

FluidFlow software is used to design and model geothermal two-phase flow systems with separator stations, injection systems as well as cooling water systems. Companies which use FluidFlow to model geothermal piping systems include Philippine Geothermal Production Company, MTL Consulting Engineers in New Zealand and EFLA Consulting Engineers in Iceland.

FluidFlow is used by EPC’s as well as owner operators to design, develop and optimise piping systems for geothermal, coal and natural gas power plants. The systems modeled include natural gas delivery systems, water and steam systems (single and two-phase), air systems, ash handling systems, chemical systems and water treatment systems. As part of this process, designers utilise the powerful automatic equipment sizing tool to size pipes, fans, pumps, control valves, safety relief valves and orifice plates. 

This discussion attempts to outline some of the basic principles of geothermal power plant systems. Further reading is recommended and the references outlined serve as a useful resource. 


  1. Pressure Drop in Large Diameter Geothermal Two-Phase Pipelines. 
  2. A Guide to Geothermal Energy and the Environment, Geothermal Energy Association. 
  3. Flow and Heat Transfer in Geothermal Systems, Aniko Toth, Elemer Bobok.
  4. Geothermal Power Plants Principles, Applications, Case Studies and Environmental Impact, Ronald DiPippo. 

Guide to Help Minimize Safety Risk when Expanding or Altering Process Plant Equipment

Increasing the capacity or yield of existing chemical process sites through expansion or upgrades is often a cost-effective means of increasing capacity without incurring the higher costs associated with embarking on a new greenfield construction. In order to complete upgrades to existing facilities successfully, the works required must be given careful design consideration with particular attention given to process engineering and plant safety. Large-scale projects may require considerable alterations to plant layouts and complex engineering alterations that can pose quite challenging considerations from a safety viewpoint. Indeed, the safety implications can often be as challenging as those posed by a greenfield site development.

Small-scale upgrades by their very description may seem minor however, they can also pose considerable safety design challenges and implications. 

In an effort to mitigate risk and reduce the incidence of serious hazards, it is crucial to examine and fully understand some common pitfalls that can contribute to major accidents at chemical process plants.  

Important Systems

Relief-valve systems: Relief valve devices are essential to maintaining plant safety and avoiding potentially devastating accidents. Unfortunately, however, proper sizing and implementation of safety relief valves is sometimes misunderstood or overlooked. When existing process plants expand, the additional throughput capacity that must be handled by the relief valves in the facility can often push the limits of the original design. To compensate for this, it may be necessary to replace existing relief valves for new devices. In some cases, valve replacement is not enough. If an existing relief valve is replaced with a new device and the existing inlet or outlet piping size and configuration is retained, it can sometimes create a bottleneck around the new device. Furthermore, the existing relief collection header piping systems may be inadequate for the new higher-pressure levels which are likely to be encountered by the upgraded system. A thorough review of the entire plant is therefore necessary to determine exactly what needs to be replaced and how it should be completed.

When carrying out upgrade works on an existing process plant, consideration should also be given to relief valve systems which are not in the immediate location of the works. Relief valve systems which are located hundreds of meters away can also be impacted by the works and therefore, should not be overlooked as part of any design. If any relief valve is inadequate or fails to meet the proper specifications, the effects could be disastrous.   

As part of any design or upgrade works, there can sometimes be a tendency to overestimate the capabilities of a particular piece of equipment. As an example, equipment such as a boiler might be capable of handling an increase in pressure as the result of a new development. However, the existing connected piping or valves attached to the boiler may not be able to handle the pressure increase. 

It is recommended that a full and complete audit of existing and any proposed new relief-valve systems should be a part of any plant modification, expansion or upgrade. It is vital that plant operators collaborate with contractors to ensure that no part of the audit is overlooked, misunderstood or rushed. Problems with safety relief valves can potentially be one of the most devastating and dangerous problems encountered at a chemical process plant.

Pipe racks: Existing piping should be thoroughly assessed and reviewed during any type of plant upgrade or expansion. In a mature plant, pipe racks can often exhibit little or no room for expansion. If the existing facility has in the past undergone modifications, it can be commonplace to discover that some of the existing pipe racks are overloaded over and above their original rated design capacity. This can pose a serious safety risk. For this reason, it is always prudent to add additional pipes to increase flow. 

Any limitations with existing pipe racks should if possible be identified early in the design process so that there are no unexpected costs for rack expansions. 

Piping: The presence ofdead-leg pipes can pose a potential hazard. These are pipes which have been disconnected from old redundant process systems and have been left behind in-situ. Dead-leg pipes can pose a safety risk and furthermore, take up unnecessary space on pipe racks. Any stagnant liquids left in the disconnected dead-leg pipes can lead to pipe corrosion or rupture in the event of the liquid freezing. 

Process plants often have dead-leg pipes that have not been part of an active process for years. As dead-legs increase in number around the site, the risk to the plant multiplies with them. When dealing with fluids such as hazardous chemicals, even small amounts of pooling fluid in dead-legs can lead to significant ruptures or leaks. Process plants should have regular management programs in place to inspect piping. While there are scenarios where it may make sense for a dead-leg to remain, any dead-legs deemed unnecessary should be carefully drained and removed from site. It is good practice to include these works as part of any plant expansion or modification projects.

Piping and pipe racks can be found everywhere in chemical process plants. It only takes a problem in one of those pipes for a safety incident to occur. For this reason, it is 

recommended that a thorough inspection of existing pipes and pipe racks is necessary as part of any planned expansion or modification. 

Access and confined-space issues: The addition of new equipment onto an existing plant can have a significant impact on plant layout. Large vessels installed to increase output can also create access/confined space hazards. In general, placement of new equipment can create unsafe confined spaces or limit access to existing critical plant infrastructure. While a certain number of confined spaces in a chemical process plant may be unavoidable, designs should try to minimize them as much as possible. 

During the construction works, there are often temporary installation requirements which can restrict access to certain areas of the existing plant. Any additional equipment that is necessary to complete a project, such as scaffolding, should be placed carefully to ensure that it does not obstruct  or restrict vital access areas. 

Even under the best of circumstances, confined spaces can be extremely dangerous. Potential safety hazards associated with confined spaces include exposure to hazardous chemicals, the ignition of flammable liquids or gases and asphyxiation.

Labelling/Identification (especially for chemicals and piping): Having proper equipment labelling and documentation is the simplest and the least costly way to improve safety. Unfortunately, it can also be one of the most overlooked ways of improving safety. Further information on labelling can be found in the ANSI/ASME A13.1 pipe-marking standard. 


This blog attempts to outline some items for consideration when expanding or upgrading an existing chemical process plant. It by no means covers all aspects of any design of such a facility but more attempts to document a few items worth considering as part of this type of development. 

A previous blog on relief valve sizing may also be of interest as it also addresses some interesting points in pressure relief devices. 



Two-Phase Gas-Liquid Flow Discussion.

Two-phase gas-liquid flow is a common phenomenon in nuclear, geothermal, chemical, oil and gas and refrigeration industries where gas-liquid mixtures are transported in piping systems. Two-phase pipe flow is the simultaneous flow of a gas (or vapor) and a liquid exhibiting different physical properties through the same pipe. A typical example of two-phase flow is a water-steam mixture. The prediction of the fluid physical properties and pressure drop in two-phase pipe flow is of vast importance during the design and operation stages in all applications to ensure the fluid does not deviate from its operational envelope and cause damage to personnel, plant and equipment. Due to the complexities of two-phase flow, it is commonplace to apply modelling techniques to predict the behaviour and characteristics of two-phase flow. 

The physical properties of the two-phase fluid are not constant or uniform throughout the piping system and flow-induced pressure drops can cause considerable changes in magnitude of one phase compared to the other. For instance, some of the liquid can evaporate (flash) due to pressure drop across pipelines or through an orifice which results in an increase in gas/vapor. FluidFlow assumes that instantaneous isenthalpic flashing occurs. The software also identifies the physical properties of the fluid at the inlet and outlet of each pipe and fittings. A common feature of two-phase flow systems is that the flowing velocities increase and the mixture density decreases. 

In cryogenic applications, two-phase flows are usually a mixture of a cryogenic liquid along with its corresponding vapor. A mixture of liquid helium and helium vapor is a typical example. An advantage of two-phase flows is that it can provide isothermal heat sinks. Due to the latent heat of boiling, heat added to the liquid portion of a two-phase flow will contribute to the conversion of the liquid to vapor in the mixture but will not increase the mixture’s temperature. This characteristic is used in cooling schemes such as those found at the Hadron Collider, HERA and Tevatron magnets. 

There are also a number of disadvantages to two-phase flows. In comparison to single-phase pipe flows, these flows typically have a higher pressure drop and flow instabilities may develop that result in pressure surges and vibrations. Incorrectly designed systems can also trap the vapor phase against surfaces resulting in inadequate cooling or energy transfer and many flow components such as pumps and flow meters may not function properly in two-phase flows. Furthermore, in cases where the flow is on an incline, the liquid may flow to the downhill side. This may therefore necessitate the use of shorter flow paths to maintain any desired liquid-vapor ratio. 

Due to these issues and complexities, detailed modelling is required whenever two-phase flows are used in cryogenic systems, to ensure proper performance. Care must also be taken that a nominally single-phase (say pure liquid) flow in a system does not suddenly become two-phase due to heat inputs or pressure drops resulting in unexpected operating problems.

Two-phase mixtures flow in pipelines in several configurations or structures called flow patterns or flow regimes. This describes the arrangement and relative fractions of the liquid and vapor in the flow. In the case of horizontal or near-horizontal pipe flows, typical flow patterns are stratified smooth, stratified-wavy, elongated-bubble, slug, annular, wavy-annular and dispersed-bubble. Some of these flow patterns are shown in Figure 1. Typical flow patterns for upward vertical pipe flow are bubble, slug, churn, annular and dispersed-bubble. 

Figure 1: Two-Phase Flow Pattern Map (FluidFlow).

The fluid flows in different flow patterns such as slug, churn and annular. Some of these flow patterns can generate dynamic fluid forces which in turn may induce structural vibration. Excessive vibration may lead to component failures which must be avoided as they can produce significant economic, environmental and even human losses. 

Therefore, the knowledge of flow conditions including flow pattern as well as flow-induced vibration (FIV) can have a significant impact on the proper design and operation of piping systems. 

Two-Phase flow; How can I solve my Liquid-Gas system

The calculation of frictional pressure loss for two-phase gas-liquid flow is a complex process. The coexistent flow of two phases complicates the theoretical and empirical approaches which are available. This means that a complete analytical solution is not possible. After 60 years of extensive research it is rare to find two correlations with exact predictions. In an effort to overcome these shortcomings the FluidFlow Two-Phase Module provides eight choices of correlation that represent some of the most successful approaches to this complex problem. 

There are three main approaches you can take to modelling two-phase gas-liquid flow systems in FluidFlow. 

  1. Define the temperature and pressure of the fluid at the inlet of the system.
  1. Define the vapor quality of the fluid at the system inlet. 
  1. Define multiple inlet boundary conditions with different fluid types. 

When using any of the above three approaches, the software will automatically establish the physical properties of the fluid(s) and automatically track the fluid phase state throughout the system. The appropriate pressure loss correlations are also applied automatically for each element in the system. 

Fluid mixtures can also be created either dynamically in the model or using the fluids in the database to create a mixture. 

Method of Solution

The pressure gradient (ΔP/L) for two-phase flow is not constant but varies along the pipe as a function of temperature and pressure. This means that the pressure drop must be calculated by integrating the pressure gradient along the pipe. The calculation approach used by FluidFlow is similar to that used in compressible flow calculations. The following steps are made:

1. A pipe increment is selected based on a small change in pressure P1 and P2. The length of this increment is not yet known.

2. Upstream temperature, pressure, quality and physical properties are determined. Physical properties for each phase and the mixture physical properties are needed. Crucial here is FluidFlow’s fluids database containing the thermophysical properties of more than 1200 fluids.

3. A flash calculation is then performed to determine the downstream quality.

4. From the downstream properties, FluidFlow determines the flow regime and then determines the incremental length of this segment. Determination of the incremental length depends on the friction loss calculation method used.

5. Steps 1 to 4 are repeated until the end of the pipe is reached. The incremental length step size therefore shortens as the calculation moves down the pipe. For the last segment, which will never be the exact length required, we use interpolating functions based on results from previous segments.

Example 1. An Air-Water Two-Phase Model (Constant Quality).

The two-phase condition can be specified at a boundary for a single fluid or we can mix gas and liquid streams in order to make a two-phase mixture. We will use the second method in the following example.

Figure 1: An Air-Water Two-Phase Model (Constant Quality).

The model in Figure 1 shows two known flows inlets (one fluid water and one fluid air) combining and being heated via a plate exchanger, then flowing to a separation vessel (5). The red dot position on the Knock Out Pot (Separator) represents the liquid outlet and the yellow dot position represents the vapour outlet. The design conditions for thee components are outlined below. 

Design Pipe Data:

  • Pipes (-1 and -2 connecting the known flows to the connector) 0.5 m in length 2″ Sch 40 pipe.
  • Pipe (-6 from the connector to the plate exchanger) 60 m in length and an inside diameter of 50.8 mm.
  • Pipe (-3 connecting the plate exchanger to the Knock Out Pot) 60 m in length and an inside diameter of 50.8 mm.
  • Pipe (-4 vapour outlet from Knock Out Pot) 5 m and 6″ Sch 40 pipe. Pipes (-5 and -7 liquid outlet from the Knock Out Pot) 10 m and 2″ Sch 40 pipe.

Calculation method – Beggs Brill.

Overview of Results:

This is an example of two-phase flow with constant quality. This means that the vapour mass fraction is constant and there is no mass transfer between the phases. It does not mean that the pressure loss per unit length is constant or that the velocity between the two phases is constant. In the first pipe section after mixing (pipe -6) the gas superficial velocity increases from the start to the end of pipe -6. For 60m of pipe -6 the total pressure loss is 146791 Pa, but the friction loss is 145349 Pa. Since the pipe is horizontal the difference is the acceleration loss. After the exchanger the mixture has experienced a temperature increase of 30 °C. The total pressure loss in the pipe after the exchanger (-3) is 208629 Pa (pipe -3 is identical in length and diameter to -6). This is because gas volume and velocity as well as other fluid properties have changed with the increase in temperature in the outlet pipe. You can get a feel for the differences by displaying the Beggs-Brill flow pattern map.

Figure 2: Beggs Brill Flow Pattern Map (FluidFlow).

Example 2. A Refrigerant system with changing quality.

In the example shown in Figure 3 we have a single fluid (R-152a) flowing through the system from a known pressure inlet (node 1) at 1.16 ATM and the fluid is at its saturation temperature of -18°C. We have specified a vapour quality (vapour mass fraction of 0.18) at the inlet and we are carrying out the calculation using the Whalley Criteria calculation setting. The Whalley Criteria allows the software to select the most appropriate correlation based on viscosity and gas / liquid ratios.

Figure 3: R-152a Two-Phase Model (Changing Quality).

Overview of Results:

This is an example of two-phase flow with changing quality. This means that the vapour mass fraction is not constant and there is mass transfer between the phases. You can see this in the results of all flowsheet elements. When we examine the vapor quality entering and leaving each node or pipe and you will see that quality is increasing as we flow through the system. This is because the pressure falls as the fluid flows along a pipe (or across a bend) and so some of the liquid boils to form additional vapour. This is known as flashing and FluidFlow assumes that instantaneous isenthalpic flashing occurs. You should also notice that velocities are increasing and that mixture densities are decreasing.

At the heat exchanger we are adding some 30 kW of heat and this has the effect of vaporising additional liquid. Across this element the quality increases from 0.19627 to 0.45889 (19.6 & 45.8% vapor).

Orifice Plates & Automatic Sizing of Orifice Plates

An orifice plate is essentially a disc with a central aperture which is inserted into a pipeline between a pair of flanges. Orifice plates are mainly used to achieve restricted or controlled flow of a process fluid. What happens when fluid flows through an orifice plate? In simple terms, when the fluid flowing in a pipeline passes through an orifice plate, the fluid velocity increases through the reduced orifice area and develops a differential pressure across the orifice device. The differential pressure created is related to the beta ratio (β) of the orifice plate where the beta ratio is the ratio of orifice bore diameter / inside pipe diameter (d/D). In general, an orifice plate will typically exhibit a bore diameter between around 30 to 75% of the pipe inside diameter. A beta ratio of 0.6 therefore exhibits that the orifice bore diameter is 60% of the pipe inside diameter. Therefore, the smaller the beta ratio, the higher the differential pressure generated. In working practice, the selected beta ratio is a generally compromise between the flow rate required and the desired differential pressure. FluidFlow software enunciates a message to the designer if the calculated beta ratio is outside of the recommended range as described in ISO 5167. 

The typical pressure profile across an orifice plate is shown in Figure 1. 

Figure 1: Typical pressure profile across an orifice plate.

In Figure 1 we can see that the static pressure upstream of the orifice plate increases slightly due to the effect of back pressure and the decreases sharply as the flow passes through the orifice plate. The pressure actually continues to drop until it reaches what’s known as the vena contracta where the velocity of flow is at its highest. Beyond the vena contracta, the static pressure starts to recover however, it never reaches the upstream value. There is therefore always a permanent pressure loss across an orifice plate. 

We have seen that when an orifice plate is fitted in a pipe which is transporting a fluid, the fluid velocity will increase and at the same time the pressure will decrease. It is relatively simple to measure this decrease in pressure which can be used to determine the velocity and consequently, flow rate. 

Energy losses across an orifice plate can be significant and it is therefore necessary to introduce a discharge coefficient (Cd) which is equal to the ratio of actual flow rate to the flow rate with no energy losses or theoretical flow rate. The discharge coefficient also considers the small difference between the cross-sectional area of the vena contracta and that of the orifice plate. 

An orifice plate has the advantage of being inexpensive and simple to use but high-pressure losses are sustained. 

We can see from Figure 1 that the measured differential pressure developed across the orifice plate is depends on the positioning of the sensing points or pressure taps. Let’s look at a selection of common tap positions.  

Corner Pressure Tappings. 

Corner taps are positioned immediately adjacent to the upstream and downstream plate faces.  They are often used in installations with pipe diameters of 2 inches or lower. Corner taps are commonly used throughout Europe. 

Flange Pressure Tappings.

Flange taps are positioned 25mm from the upstream face of the orifice plate and 25.4mm from the downstream face. They are often used for pipe sizes of 2 inches or greater. Flange taps are commonly used in the U.S. 

Vena Contracta Pressure Tappings.

These taps are positioned 1D upstream and at the point of minimum pressure downstream (the vena contracta). This point varies with the beta ratio and is generally only used in plant measurement where flows are relatively constant and plates are not changed. The vena contracta taps provide the maximum pressure differential, but also the most noise. 

D and D/2 Pressure Tappings.

The position of the upstream pressure tapping is nominally equal to D and the position of the downstream pressure tapping is nominally equal to D/2 where, D is the inside diameter of the pipe.

Pipe Pressure Tappings.

These taps are positioned 2.5D upstream and 8D downstream of the orifice (position of maximum pressure recovery). 

The example case shown in Figure 2 is based on an orifice plate installed in a pipeline transporting water. The calculation has been completed based on the ISO 5167 Standard. A simple model was created in FluidFlow software using the ISO 5167 relationships. The software results were then compared with those of the hand calculation. We can expect very subtle differences between the software result and hand calculation owing to the level of comparable accuracy and the slight difference in fluid physical properties, i.e. the software being much more accurate in both cases. 

The design pipe diameter was 300mm with an orifice size of 120mm. The benchmark hand calculation was based on a fluid density of 1000 kg/m3 and viscosity of 1.2927 centipoise. The hand calculation also quotes a velocity of 2.46 m/s, a flow rate of 0.0279 m3/s and a flange tap loss of 8200 Pa. 

Figure 2: Modelling of an orifice plate to ISO 5167 Standard – FluidFlow.

Based on the available system data, FluidFlow software calculated a flange tap loss of 8170.1 Pa with an associated flow rate of 0.0279 m3/s and a velocity of 2.47 m/s. It can therefore be seen that the results compare very well. The subtle differences in results can be attributed to the differences in fluid density 999.06 kg/m3, viscosity (1.138 centipoise) etc. 

Orifice plate apertures can be configured to handle various flow measurement applications. The design of the orifice plate, its position and the position of the tappings are described and specified by BS and ISO standards to minimise pressure losses and to give reproducible results. 

Let’s consider a few typical orifice plate configurations. 

Concentric Orifice Plate.

This is generally the most commonly used type of orifice plate with the orifice bore positioned along the pipe centreline. When designing and configuring orifice plates, there are several factors which must be followed to ensure accurate and reliable measurement. The orifice upstream edge should be sharp and square. The minimum plate thickness is standardized based on orifice bore, inside pipe diameter etc. The beta ratio (d/D) should be within recommended limits to ensure conformance with recommended practices. The plate surface flatness should also be within the acceptable tolerance limits. 

Eccentric Orifice Plate.

The eccentric orifice plate is similar to the concentric device in that it also exhibits a circular opening/bore. The bore position however is offset from the centreline of the pipe. When the bore is offset and oriented downwards of the pipe centreline, this enables flow of impurities in the measured fluid through the orifice without forming deposits and sediments in front of the aperture. 

Eccentric orifice plates are often used in pipelines transporting heterogeneous mixtures, fluids carrying small amounts of non-abrasive solids or gases with small amounts of liquid or where there is the potential for formation of sediment, tar deposits etc.  

Segmental Orifice Plate.

This device is essentially a thin plate with an opening in the shape of a circle segment which is comparable to a partially opened gate valve. This device is generally used for measuring liquids and gases carrying non-abrasive impurities including light slurries or exceptionally dirty gases. 

The predictable accuracy of both segmental and eccentric orifice plates is not as high or reliable as the concentric orifice plate device. 

Automatic Sizing of Orifice Plates.

FluidFlow software gives you the ability to model specific orifice plate sizes however, the software also includes powerful equipment sizing technology which allows you to size pipes, pumps, fans, control valves, relief valves and of course, orifice plates. Previous blog-posts discuss sizing of control valves and relief valves in further detail. You can view these on the “Blog” page of our website. Let’s get back on track and consider auto-sizing of orifice plates in a simple cooling circuit with several plate heat exchangers. 

It is intended to balance the flow to each of the five heat exchangers using orifice plates. The total system flow rate is 11.96 kg/s and for the purposes of this exercise, the design temperature change across each heat exchanger is 5oC. 

The model was created and solved without any flow balancing. The result of this solution is shown in Figure 3. We can clearly see the mass flow through each heat exchanger is quite different and unbalanced. As such, the calculated temperature change across each heat exchanger varies depending on the flow rate.  

Figure 3: Unbalanced Cooling Circuit Model – FluidFlow.

FluidFlow software allows us to auto-size orifice plates based on either a design flow rate or design pressure drop. We can use the equation Q = m.C.∆T to determine the design mass flow rate (m) through each heat exchanger. The mass flow is therefore: m = Q / C.∆T = 50 / 4.18.5 = 2.392 kg/s. Using the software auto-sizing option, we can define the design flow rate of 2.392 kg/s across each orifice plate and allow the software to determine the required size of each orifice plate in the system. 

Figure 4 shows each orifice plate has been auto-sized and the calculated size required shown in each case. For instance, we would need an orifice plate with a size of approximately 13.76mm in the line serving HE1. 

We can also see the temperature change across each heat exchanger is now 5oC which matches the system design conditions

Figure 4: Balanced Cooling Circuit. Orifice plates automatically sized based on design flow rate – FluidFlow.

We can now update the model and instead of auto-sizing the orifice plates, enter the actual diameter in each case and solve the system on that basis. Figure 5 shows the results of this case. 

Figure 5: Balanced Cooling Circuit. Orifice plate sizes defined based on results calculated in Figure 4 – FluidFlow.

Figure 5 also shows the calculated corner, flange and D and D/2 tap losses for each orifice plate. 

Orifice plates can be found in most process plants owing to their simplicity and comparatively low cost. It is worth noting that, a number of factors can affect the accuracy of the device by as much as 15 % or even higher. These factors include improper installation, blockage or partial blockage, misplacement of the instrument, installation of slightly warped orifice plate, deposit build-up on plate surface, deposit build-up on pipe on either side of plate, different or unexpected flow conditions, i.e. single phase vs two phase flow.  Damaged plates can also result in a lower level of accuracy. 

This blog gives a broad overview on orifice plates as well as sizing of orifice plates for a “canned” closed loop cooling circuit example. If you are interested in reading about closed loops systems, sizing of control valve or relief valves, take a look at some of our other blog-posts.


  1. Introduction to Food Process Engineering by P. G. Smith, Peter Geoffrey Smith.
  2. Fundamentals of Orifice Meter Measurement, Emerson.
  3. ISO 5167.

Sanding of Slurry Pipelines

Sanding of Slurry Pipelines: The obstruction of flow within a slurry pipeline is sometimes referred to as sanding. Sanding can occur when particles or solids accumulate in a pipeline which in turn reduces the pipe cross section area or flow path. This is associated with an increase in friction loss. An accumulation of sanding can result in lower flow rates, lower delivered solids content and higher discharge pressures. This can negatively affect output and if overlooked or ignored, can have significant safety implications. Continuous accumulation of solids the pipeline can eventually lead to a blocked pipe.

The reduction in pipe cross section area from sanding results in a lot of energy being transferred to a small volume in the system pump. Over time, the carrier fluid or liquid will boil and the resultant steam can become pressurized. Blockages can then result in catastrophic failure where the pump may even explode. 

In some cases, a bed of particles or sediment may form on the bottom of a pipeline without completely blocking the flow path. Under this operation, the pipe continues to deliver solids however, it is often recommended to avoid operating under these type of sanding conditions. Left unresolved, this can lead to eventual pipe blockage which will require a complete shut-down of the slurry system. Continuously operating under these conditions results in higher pressure losses and higher energy requirements and associated costs.  

If you have identified sanding in your slurry lines, what options are available to you? Firstly, you can consider increasing throughout to clear the lines by increasing the pump speed. 

Secondly, you can consider introducing fine particles as these fine particles help to increase the fluid density and viscosity and may clear the solids build-up/bed on the base of the pipe. 

Another option to consider is flushing out the pipelines with water however, caution must be exercised with this option as, the pressure the pump develops when pumping water is generally lower than that it develops when pumping slurry for the same given pump speed. This therefore may or may not be a viable solution depending on the characteristics of the system. 

Of course, you can also shut down the pipeline and manually remove the blockage. This however may not be the most economical solution as it can delay site productivity and effect potential revenue. 

Selecting an appropriate software tool is a logical step to take when designing your slurry pipelines. This will help you identify operating conditions and the risk of sanding. FluidFlow helps you characterise your slurry as well as model the slurry pipeline. The software allows you to determine the deposition velocity of the slurry and you can now choose from four correlations for this function. System pressure losses are calculated using any of five methods and solids derating of end suction centrifugal pumps can also be considered. Warning messages are enunciated to help you identify any trouble-spots or risk of sanding/pipe blockage. 

FluidFlow software has been used successfully to model slurry flow systems by Worley Parsons, JACOBS, HATCH, Emirates Global Auminium, Rusal Aughinish, BHP Billiton, Ambatovy to name a few.  


  1. KSB Pumps.  

Relief Valve Sizing

Flare Relief Vent System

Relief Valve Sizing – Pressure relief devices (PRD’s) are widely and effectively used to protect process equipment such as piping systems, pressure vessels, distillation columns and other equipment from pressures exceeding the design-pressure rating by more than a fixed predetermined amount. The aim of pressure relief valves is to prevent damage to equipment, prevent injury to personnel and to avoid potential risks to the environment.

PRD’s operate at a designated pressure, ejecting mass from the process which contains energy and the release of this energy reduces the process pressure. It is therefore critical that proper relief-valve sizing is carried out to ensure the fluid has sufficient flow area to exit the process thus safeguarding the safe operation of the process plant and equipment. It doesn’t end there however. As a safety pressure relief valve must be capable of operating at all times, in addition to proper sizing, it is also essential that due care has been given to the proper selection, manufacturing, assembly, testing, installation and maintenance of PRD’s. This is essential to ensure maximum protection in line with the design intent.

There are many different ways in which high pressure can be encountered in a system such as, failure of a control valve, external fire, thermal expansion of a liquid, or a reaction which becomes out of control. Each potential cause of overpressure is generally referred to as a scenario and all potential overpressure scenarios must be identified, characterised (as credible or non-credible) and documented prior to sizing a relief valve or relief system.

Pressure relief valves have been known to be sized and selected simply by matching the connection size of an available vessel nozzle or the size of the pipeline connection. Correct relief valve sizing is a complex process and there are many methods and tools available to size such devices. Two common Standards used for the sizing of pressure relief valves are ISO 4126 & American Petroleum Institute API RP520 Part I. The latter Standard is the most widely used for relief valve sizing. These standards outline several equations for sizing relief valves for various flow conditions such as liquid flow, gas flow (critical and sub-critical), steam flow and two-phase flow.

Before we consider the subject of relief valve sizing any further, we should firstly understand some of the terminology used.

Maximum Allowable Working Pressure (MAWP): This is the maximum gauge pressure permissible at the top of a completed vessel in its normal operating position at the designated coincident temperature specified for that pressure. The pressure is the least of the values for the internal or external pressure as determined by the vessel design rules for each element of the vessel using actual nominal thickness, exclusive of additional metal thickness allowed for corrosion and loadings other than pressure. The MAWP is the basis for the pressure setting of the pressure-relief devices that protect the vessel. The MAWP is normally greater than the design pressure but can be equal to the design pressure when the design rules are used only to calculate the minimum thickness for each element and calculations are not made to determine the value of the MAWP.

Set Pressure: This represents the pressure at which the device operates. Note, small amounts of leakage may occur at 92 to 95% of the set pressure for spring-operated relief valves.

Overpressure: This represents the pressure increase over the device set pressure, usually expressed as a percentage of the set pressure. Sufficient overpressure is required to achieve full lift.

Accumulation: This represents the pressure increase above the MAWP, usually expressed as a percentage of the MAWP. When the relief valve is at the MAWP, the overpressure and accumulation are equal.

Backpressure: This represents the pressure downstream of the relief device. This includes the superimposed backpressure and built-up backpressure due to fluid discharge from the relief device through the downstream piping.

It has already been noted that PRD’S protect equipment from pressures exceeding the design-pressure rating by more than a fixed predetermined amount. The ASME Boiler & Pressure Vessel Code Section VIII outlines design requirements for pressure vessels and the relief valves protecting them as a percentage of the maximum allowable working pressure (MAWP). Figure 1 provides an overview of the ASME recommendations.  

The left-hand column outlines requirements for pressure vessels and the right-hand column outlines requirements for the relief valves. The entire figure is relative to the MAWP which is assigned an arbitrary value of 100%.

Pressure Relationships for PRVS

Figure 1: Pressure Relationships for PRVS.

Figure 1 shows that the vessel’s MAWP is equal to the relief valves maximum allowable set pressure. This relationship applies to vessels protected by a single device. If the vessel is protected by multiple devices, then one relief valve must be set no higher than the MAWP however the others can be set as high as 105% of the MAWP.

Relief valves are typically set to open at the MAWP. The anticipated operating pressure should be sufficiently low to prevent activation of the relief device during normal operation. The difference between the set pressure and the maximum operating pressure is known as the operating margin. The operating margin required depends on the type of relief device and the pressure control capability of the process.

Figure 1 also illustrates that, the allowable accumulation for pressure vessels protected by a single device is 110%. The exception to this case is fire exposure scenarios where the allowable accumulation is 121% of the MAWP. In scenarios where multiple relief devices are used for non-fire conditions, the allowable accumulation is 116%.  

Relief valve sizing is a complex multi-step process. The following steps outline some of the considerations to be given as part of the sizing procedure.  

  1. Define the Protected System:  This can include several equipment items and can become quite complex for processes involving connected equipment with different pressure ratings. An example of such a scenario would be a relief device on the top of a distillation column. The device may protect the column, condenser, reboiler and accumulator.
  2. Position the Relief Devices: Establish where overpressure protection may be required. The following bullet points outline some locations where relief devices may be required:
  • Pressure vessels require overpressure protection.
  • Piping at risk of overpressure due to process control failure.
  • Compressors, PD Pumps and turbines require pressure safety relief devices on the discharge side for deadhead protection.
  • Relief devices should be fitted to liquid-filled lines where there is risk of overpressure due to thermal expansion.  
  • Low pressure storage tanks require pressure and vacuum relief for normal operation. Tanks must also be protected from emergency events that could create abnormally high venting loads.
  • A vessel jacket requires its own overpressure protection.

3. Define the Overpressure Scenarios: There are typically multiple scenarios that may result in overpressure at each individual piece of process equipment. The plant P&ID’s should be reviewed in detail to identify these scenarios. Care should be taken at design-stage to ensure that you are working off the most up to date drawings. When all scenarios are identified, the next step is to characterise which are credible and non-credible. It is worth noting at this point that ASME Boiler & Pressure Vessel Code Section VIII requires that all pressure vessels are provided with overpressure protection regardless of whether or not there are any overpressure scenarios. This therefore requires the installation of a relief device to ensure compliance with the standard. Care must therefore be given to ensure relief protection is provided to the system and/or equipment items to ensure compliance with relevant standards.

4. Choose the type of Relief Device: As the relief valve sizing calculations depend on the type of relief valve selected, we need to select an appropriate relief valve for the given application. The most common devices are spring-operated, balanced bellows, rupture discs, pilot-operated and rupture-pin relief devices. A combination of these devices may also be used. Reclosing relief devices should be considered from both a safety and reliability perspective.

5. Obtain Design Data for Relief Valve Sizing: The design data typically includes fluid physical properties such as density, viscosity, heat capacity, heat of vaporisation and vapor pressure as well as flow rate, overpressure, back pressure and protected equipment MAWP.   

6. Determine Flow Conditions (Single or Two-Phase): At this stage it is useful to identify if the flowing fluid will be in the single or two-phase state as this also has consequences for the selected relief valve sizing equations. FluidFlow removes the need for the designer to make this qualification as the software identifies the fluid physical properties including phase state and automatically applies the appropriate equations.

Let’s consider a number of examples relief valve sizing cases.

Example Case 1: Hydrocarbon Mixture.

Figure 2 provides an illustration of a hydrocarbon mixture (butane & Pentane). This example is based on a hand calculation detailed in API RP520 Part 1. The design flow rate is given as 53,500 lb/h, the relieving temperature and set pressure is 348 Kelvin and 75 psi g respectively and the back pressure is given as 14.7 psi a (or 1 atm). The permitted accumulation is 10 % and the relieving pressure is 670 kPa absolute. The hand calculation detailed in API RP520 Part 1 notes that a discharge coefficient of 0.975 has been used and the calculated relief orifice size is given as 3179 mm2.

Relief Valve Sizing

Figure 2: Relief Valve Sizing for Butane & Pentane Hydrocarbon Mixture.

The results developed by FluidFlow software reveal a calculated size at MAWP of 3188.5 mm2 which compares well with the hand calculation of 3179 mm2. The size is slightly different which is expected and this can be attributed to a number of factors:

  1. The physical properties (molecular weight etc) of our gas mixture is slightly different to that used in the hand calculation.
  2. FluidFlow does not assume gas ideality but solves for real gas conditions using an equation of state (three EOS to choose from).
  3. The hand calculation neglects the effect of connected piping whereas FluidFlow solves the system and considers the effects of any connected pipes.

Note that the result for Calculated Size is always greater than the Calculated Size at the MAWP (Maximum Allowable Working Pressure) due to pipeline and entry/exit losses. When using the API pressure loss model, FluidFlow suggests the closest matching standard size orifice available which can then be modeled in your system.

Example Case 2: Atmospheric Distillation Column.

The next sizing and selection case (Figure 3) is for an atmospheric distillation column of a crude oil refining process (Ref 5). The design flow rate of the gas is given in the publication as 50,000 kg/hr, molecular weight is 94 kg/kmol, the relieving temperature and pressure is 473 Kelvin and 4.725 barg respectively and the back pressure is given as 14.7 psi a (or 1 atm). The permitted accumulation is 16 % and the set pressure is 3.2 barg. The publication includes a hand calculation which notes that a discharge coefficient of 0.975 has been used and the calculated relief orifice size is given as 93.17 cm2.

It should be noted that the hand calculation includes simplifying assumptions of gas ideality and neglects any connected piping. The scenario was modeled in FluidFlow which solves for real gas conditions and includes the effects of any connected piping. The orifice size calculated based on API is 90.37 cm2 as outlined in Figure 3.

Relief Valve Sizing

Figure 3: Relief Valve Sizing for Atmospheric Distillation Column of Crude Oil Refining Process.

The API Standard outlines Standard Orifice Sizes and as shown in Figure 3, FluidFlow has suggested the next closest size match is orifice size “R” (103 cm2). This matches the hand calculation. A Sarasin RSBD valve model for “R” orifice (size 103 cm2) was selected (refer to the bottom example in Figure 3).

The hand calculation applies a correction to establish the corrected flow rate based on orifice size “R” being used (103 / 93.17 * 50,000 = 55,275 kg/hr). An approximation to the FluidFlow results can also be established by a hand calculation as follows; 103 / 90.37 * 50,000 = 56,987 kg/hr. Figure 3 shows that when using orifice size “R”, the actual corrected flow rate established by FluidFlow is 56,984 kg/hr.

Example Case 3: Supercritical Butane.Supercritical fluids exhibit characteristics typical of both liquid and vapors and their physical properties can be strong functions of pressure and temperature and may deviate appreciably from ideal gas behaviour. The following is a sizing example for supercritical butane which for the purposes of this case, uses the API standard sizing equations. Note, this approach is based on air, the ideal gas law, an uninsulated vessel with no mass and no change in fluid temperature. CEP Magazine 2002 includes a hand calculation of this case and arrives at an orifice size of 0.034 in2. The design conditions are outlined in Figure 4. The relief valve set pressure is 800 psig with a 21% allowable overpressure for the fire case.

Relief Valve Sizing

Figure 4: Relief Valve Sizing for Supercritical Butane.

FluidFlow has solved the case arriving at an orifice size of 0.040 in2 at MAWP which considering the software takes into account connected piping, real gas conditions etc, is a close match.

It is worth re-iterating that the sizing approach outlined in the API Standard for supercritical flow conditions is generally based on ideal gas and incompressible fluid behaviour and as such, may not be appropriate for specific cases. It may therefore lead to conservatively large orifice areas. FluidFlow software on the other hand doesn’t make these simplifying assumptions and solves for real gas conditions using an equation of state thereby offering a much more accurate solution.  

Example Case 4: Flare Relief Vent System.

The final example is of an extractive distillation plant which has 12 safety relief valves. There are two major relief scenarios: cooling water failure and external fire. The governing case in this instance is the cooling water failure as it occurs plant wide. External fire occurs only at localised areas and the relief loads come from just a small number of relief valves. This model therefore considers only the cooling water failure case. For simplicity, a set pressure of 5 kg/cm2 g has been assumed for all safety relief valves. This solution is based on an example case outlined in Chemical Engineering Magazine.

The relief rates are shown in Figure 5. The system is made up of a total of 395 M of pipework ranging in diameter from 4 to 12 inches.

Flare Relief Vent System

Figure 5: Flare Relief Vent System.

Based on the worked example in the hand calculation, the point B has a pressure of 1.6234 kg/cm2 a and point C a pressure of 1.9616 kg/cm2 a. This compares well with FluidFlow which has established a pressure of 1.62823 kg/cm2 a at point B a pressure of 1.9569 kg/cm2 at point C (see Figure 5). The total flow rate in this final line based on 7 of the 12 relief valves venting is 95,648 kg/hr.

Relief Systems Installation Tips:

  • Safety relief valves should be connected to the vapor space of the protected equipment.
  • To ensure reliable overpressure protection, it is best to install relief valves without any isolation valves. However, relief valves occasionally do no re-seat properly and start to leak. Therefore, in some cases, two safety relief valves are installed to allow replacement of relief valves that are leaking while the plant is in operation. This also facilitates regular testing and servicing of relief valves without interrupting plant operations. In scenarios where two relief valves are installed, isolation valves are provided for each relief device. This is to facilitate isolation of valve “A” for maintenance while bringing valve “B” online when required. In these cases, a ¾ inch bleeder valve with an isolation valve is recommended in between the isolation valve and the pressure relief valve. This is required because when valve “A” is isolated, the section between the isolation valve and the relief valve is still at the operating pressure of the column. The bleeder is therefore used to depressurise the system locally before the valve is taken offline.
  • The inlet line should be self-draining back to the process vessel. This prevents accumulation of liquid that can corrode or block the system. Likewise, the outlet line from the pressure relief valve should be self-draining to the flare header. To meet this requirement, it is recommended the relief valves installed at high points in the system.
  • When two relief valves are provided, it is mandatory to provide a mechanical interlock system between the respective isolation valves to ensure that one isolation valve is open at all times. This eliminates the potential risk of operator error, i.e. a scenario where both isolation valves are in the fully closed position rendering both relief valves unavailable to offer protection of the equipment.  
  • Whenever a bursting disk is installed upstream of a relief valve it is important to have a pressure indicator in the section between the two items. This is in case a pin hole leak occurs in the disk which would allow vapors to pass through to the section between the disk and the relief valve. Over time, the pressure up and downstream of the disk would reach equilibrium and the disk would never rupture.
  • Relief valves in water, air or steam service are connected to the atmosphere through a short section of pipework. To keep this pipe free from liquid accumulation, a small weep hole is drilled at the lowest point of the pipe.


Economics also play a part in the design of safety relief systems. For instance, occasionally process industries use exotic materials, such as titanium and Hastelloy C. In such cases, instead of having a pressure relief valve made of Hastelloy C, it may be more cost effective to have a rupture disk made of Hastelloy C followed by a stainless-steel pressure relief valve.

In order to provide adequate protection of life and property, a safety pressure relief valve must be available and capable of operating at all times. It is therefore vital that the device has been carefully sized and selected. This article attempts to outline some of the main considerations when sizing and selecting safety relief valves. Further reading is recommended on the topic with the references outlined below serving as a useful starting point.  


  1. Sizing Calculations for Pressure-Relief Valves, Chemical Engineering Magazine.
  2. Sizing Pressure-Relief Devices, CEP Magazine.  
  3. Rigorously Size Relief Valves for Supercritical Fluids, CEP Magazine.  
  4. American Petroleum Institute, API RP520 Part 1.
  5. Sizing & Selection of Pressure Relief Safety Valves, Eastern Refinery Ltd.
  6. ASME Boiler & Pressure Vessel Code Section VIII.

To explore this subject further, please see a webinar recording below giving an overview of how FluidFlow can be used to automatically size your pressure relief devices to API & ISO standards:

Mine Tailings Transportation & Disposal

Mining of valuable minerals or metals usually involves the removal of vast quantities of earth from the ground in order to obtain a relatively small amount of the desired material. The material left over after the refining process is referred to as mine tailings. 

Mine tailings typically consist of process effluent and finely-ground rock generated at the mine processing plant. Tailings have no financial gain to a mine or mineral operator and as such, tailings are usually stored in the most cost-effective way possible to meet site specific factors as well as adhere to relevant environmental regulations. 

Tailings can be quite abrasive and often corrosive. The selected pipelines and equipment must therefore be able to adequately withstand the effects of the tailings. Valves used in tailings transportation systems are often selected such that they feature an enclosed operating mechanism to ensure the valve will be operational when operated.  

The physical properties of the tailings, tonnage and local site topography often dictate the type of transportation system required to manage the tailings both cost effectively and efficiently. Tailings are sometimes used as backfill in underground mines, stored in open pits, dried and stacked, or pumped through pipelines into surface tailings ponds on site to allow the solid particles to separate from the water and settle on the bottom. The often-remote location of tailings ponds means that reliability, safety and low maintenance are vital since access is not always straightforward or convenient. 

Increasingly strict environmental regulations as well as challenging economic conditions demand tailings piping systems that transport tailings at higher concentrations, travel over longer distances and are designed to have a longer lifespan. 

The tailings slurry consistency is dependent upon the mass percentage of solids, particle size distribution (PSD), type of tailings, specific gravity (SG) as well as the extent of thickening at the processing plant. Tailings are often pumped at a solid content of 25 % for low specific gravity coal tailings to over 50 % for hard rock metalliferous tailings. 

Prior to the transfer of tailings to a pumped storage facility, tailings often undergo a thickening process. This thickening step facilitates the recycling and reuse of the process water which is transferred back to the mineral processing plant. This reduces water losses and reduces the overall water demand at the processing plant. 

The ability of a pumped piping system to transport the tailings slurry is a function of the slurry rheology. In order to determine if tailings can be pumped with existing pumping equipment, the viscosity and yield stress need to be determined. The higher the slurry yield stress, the more difficult it is to pump the slurry (for a given pump type). The yield stress is strongly dependent on the solid content of the slurry. 

The design of pumped tailings pipelines can be quite complex owing to the variability of tailings properties and flow rate for the mining operation.  A system must therefore be designed to allow for this variation in operating conditions. The design of these systems, usually over large distances with different terrains and elevations can become quite complicated and extremely difficult to design accurately by hand or using a spreadsheet. 

The use of centrifugal pumps is still quite a common method of pumping tailings in piping systems. Many of our customers use FluidFlow to automatically size slurry pumps which can be costly plant items. As such, it is important to ensure the pump sizing and selection process has been completed accurately and that the pump(s) are operating at a safe, reliable and efficient duty point. 

The use of an appropriate software tool such as FluidFlow allows engineers to design, model and optimise tailings transportation pipelines and consider variability of tailings concentrations and properties. Any risk of saltation or pipe blockage is identified helping the designer develop a safe, reliable and effective system. Solids derating of centrifugal pumps can also be considered and evaluated. 

FluidFlow has been used successfully in the mining and metals industry for many years to design and model slurry transportation systems as well as mine dewatering systems, mine venting systems to name a few. Figure 1 

Modeled Tailings System - FluidFlow.

Figure 1: Modeled Tailings System – FluidFlow.

Some of our users include BHP Billiton, Rio Tinto, Rusal Aughinish, Aurecon, WorleyParsons, Emirates Global Aluminium (EGA), Ambatovy and JACOBS. The Ambatovy site in Madagascar features a pumped tailings pipeline which transfers the tailings over a distance of 15km to a 750 ha storage facility. 


  1. Paterson and Cooke.

FluidFlow v3.45

General Release summary: This release adds a new node kind, color coding, and access speed over a LAN improvements. This release also contains bug fixes.


• Gas Regulator node added.
• Pipe Color Coding for Flow, Pressure, and Velocity.
• Ability to load from and save to JSON format.
• Ability to clone, load, and save ‘Database’ records.
• Read-only database access for improved calculation speed (over LAN).


• Casson calculation improved.
• Text on flowsheet now snaps to grid.
• Consolidation of Excel and Tools menu items.


• Control Valves fixes.
• Composite Chart fixes.

Control Valve Selection

A previous blog-post discussed the importance of control valve sizing and energy optimization opportunities. This blog-post will focus more on the topic of control valve selection although, both topics shouldn’t be considered to be mutually exclusive. 

When selecting a control valve for process plant, there are many things to be considered. These can include the valve flow characteristic, size, valve body and trim materials, noise, potential for damage from cavitation or flashing, actuator type and size, dynamic response to changes in control signal etc. This summarises the typical considerations when making a control valve selection. 

Selecting an improperly sized control valve can have serious consequences on safety, operation and productivity. The following list outlines some of the things to consider when making a control valve selection:  

  1. Give careful consideration to selecting the correct materials of construction. Take into consideration the components of the valve that come in to contact with the process fluid such as the valve body, the valve seat or any other valve components exposed to the process fluid. 
  2. Consider the operating temperature and pressure the control valve will be exposed to. Consider the local ambient atmosphere and any corrosives that can occur which may affect the exterior of the valve. 
  3. Consider the degree of control you require and ensure the selected valve is mechanically capable of achieving the desired operating conditions. 
  4. Consider the inherent flow characteristic of the control valve you are selecting. Different valve types have different flow characteristics. The flow characteristic can be generally thought of as the change in rate of flow in relationship to a change in valve position. This item is discussed in a little more detail later. 
  5. Aim for optimal valve travel. When a valve is sized correctly, the range of operation will correspond well to the control range of the valve. Some industry literature recommends travel at normal flow should fall within 50 to 70 percent opening angle. Travel at maximum flow should fall below 90 percent whereas travel for minimum flows should be above 20 percent open to avoid erosion of the trim. When modelling a control valve in FluidFlow, the software enunciates a warning message if the valve position falls outside of an optimal operating range. Users can also adjust the desired settings for minimum and maximum valve positions. This helps prompt the designer into considering valve position for the given design operating conditions. Ultimately, it also helps the designer make a better and more appropriate valve selection for the application in hand. 
  6. Avoid oversizing a control valve. If the control valve is too large for the required application, only a small percentage of travel is required. This is due to only a small change in valve position having a large effect on flow which in turn makes the valve hunt. This can cause excessive wear. Some published literature sources recommend sizing a control valve at about 70% to 90% of travel. 

This list represents just some of the criteria to be considered when selecting a control valve. It is generally recommended that the final valve selection is discussed with an appropriate and experienced supplier or manufacturer before making your final selection. 

Control Valve Flow Characteristic

The flow characteristic of a valve represents the inherent relationship between the valve opening and flow rate. As a valve gradually opens, the flow characteristic allows a certain amount of flow though the valve at a particular opening percentage. This permits predictable flow regulation through the valve. The most common flow characteristics are linear, quick opening and equal percentage. 

Linear Flow Characteristic

This flow characteristic exhibits a linear relationship between valve position and flow rate. The flow through the valve varies directly with valve stem position. 

Linear Flow Characteristic – FluidFlow.

Quick Opening Flow Characteristic

The flow characteristic of a quick opening valve is such that for a relatively small initial change in valve stem travel, a large increase in flow occurs. The noticeable characteristic of this valve type is that maximum flow is achieved at a relatively low percentage of the valve stem range. 

Quick Opening Flow Characteristic – FluidFlow.

Equal Percentage Flow Characteristic

The flow characteristic of an equal percentage valve produces equal percentage changes in the existing flow for equal increments of valve travel. The change in flow rate is proportional to the flow rate just before the change in position is made. 

Equal Percentage Flow Characteristic – FluidFlow.

The above summarises the most common valve flow characteristics.  

Control valves actually have two characteristics, inherent and installed characteristic. The inherent characteristic is that published by a valve manufacturer based on tests conducted on a system where care is taken to ensure the pressure drop across the valve is held constant at all valve opening positions and flow rates. The inherent characteristic therefore represents the valve flow capacity and valve opening position when there are no system effects involved.

The installed characteristic is the relationship between the valve position and flow in the system taking into account any changes in the pressure differential available to the control valve due to the flow squared relationship between flow and piping pressure losses and/or the behaviour of a centrifugal pump’s head curve. 

The performance of control valves in a process system can have a dramatic effect on the plant efficiency, asset life cycle costs and overall profitability. It therefore goes without saying that the cost-effective operation of any plant, industrial or otherwise, requires considered design and careful control valve sizing and selection. A correctly sized control valve can provide significant savings as well as increase process availability, reduce process variability and reduce maintenance costs. Correctly sized control valves also last longer than unmatched or incorrectly sized valves. 

Oversized valves have a higher capital cost and tend to cause instability in the operation of the system whereas undersized valves simply won’t pass the required flow of fluid in the line. 

As designers, it is therefore worth giving careful consideration to both the sizing and selection of the control valve to affect efficient and effective operation of a process plant whilst optimising operating costs. 

Selecting the Right Device

The essential steps required to optimize control valve performance as well as prevent erosion problems include proper valve sizing and selection of valve body and trim materials. They could mean the difference between continued operation and unplanned shutdowns. There are of course other decisions involved in selecting the right valve solution. Many companies choose globe type valves for their proven performance and life cycle advantages. When compared to other available valve designs, this valve offers:

  • Better control performance.
  • Better low or partial load performance. 
  • High differential pressure across the valve. 
  • Smaller physical profile than a comparable ball valve. 
  • Use for steam, water or water/glycol fluids. 

In general, a globe valve modulates flow through movement of a valve plug in relation to the ports located within the valve body. The plug is connected to the valve stem which in turn (no pun intended !!!) is connected to the actuator. 

Importance of Trim Material

Proper control valve selection can result in a high level of performance but how can this be maintained? Like other piping components, control valves can wear over time which can produce continued deterioration of the initial control valve performance. Left unchecked, this progressive deterioration can eventually lead to failure, shutdowns we well as the associated repair costs and financial impact of equipment shutdowns. 

Trim refers to the internal elements of a control valve and these elements are a crucial consideration in the process of valve selection. Trim typically includes the valve seat, disc and stem as well as the sleeves within the valve which are required to guide the stem. The interface between the disc and seat along with the relation of the disc position to the seat normally determines the performance of the control valve. 

A control valve’s trim may be selected to create a variety of passage shapes that control the flow in specific ways. The gap within the valve opens by moving the plug, disc or valve away from the seat. The length of the valve stroke determines the opening size and how much fluid passes the seat. Changing the size of the internal gap can increase, decrease or maintain the flow though the valve. Whenever the process parameter or variable being controlled does not equal the design requirement, the control valve operates and alters the opening to achieve the setpoint conditions. 

Manufacturing plants can encounter significant problems from erosion or weakening of valve bodies or trim components from severe process conditions. Typical damage can include seal rings and gasket loss, stem, body and trim retainer wear on the seat ledge, plug, seat ring and cage wear and packing leakage. 

There are several common causes for premature trim wear in control valves. One example would be where flashing occurs, i.e., when the pressure of the flowing fluid falls below its vapor pressure and changes the fluid phase-state from a liquid to a vapor. Small vapor cavities are formed under these conditions which cause wear at the outlet of the valve and its trim components. 

Cavitation is similar to flashing except the fluid pressure recovers to a level above its vapor pressure at flowing conditions. This causes the vapor cavities to implode producing impinging jets with the potential cause severe erosive damage. Outgassing occurs when the pressure of a fluid drops below the saturation pressure of a dissolved gas. When this point is reached, the gas separates from the fluid or solution and produces a high velocity erosive vapor droplets. The simplest way of appreciating this occurrence is to think of an unopened can of soda/soft drink/fizzy pop.  Once we open the can, which of course is under pressure, the sudden pressure drop causes some of the carbon dioxide to escape from the solution as a gas. When the outgassing condition arises in a flow system, in addition to the wear from vapor droplets, it can lead to vibration and eventually the trim can no longer shutoff the flow or maintain the desired flow stability. 

Benefits to plant operators 

Demanding business or manufacturing environments require the most accurate and reliable control of production processes possible. The failure to meet and achieve specific operating standards can produce an inherently inefficient plant, can lead to serious consequences for quality and safety and can significantly affect the financial margins for the final product. Optimum control valve performance is therefore vital in preventing such scenarios. 

Industrial organisations can benefit greatly from working closely with their manufacturer representatives or instrumentation suppliers to initially specify appropriate measurement and control devices. This collaboration can achieve important performance criteria including:

  • Precise flow and pressure control. This produces stable and consistent production results. 
  • Efficient energy usage.
  • Reduced operating costs.
  • Fewer unplanned and undesirable plant shutdowns.
  • Increased plant availability.
  • Lower maintenance and repair costs resulting in longer valve trim life. 

Control valves are required to withstand the erosive effects of the flowing fluid while maintaining an accurate position to control the process. In order to successfully perform these tasks, control valves need to be sized accurately and correctly for the application as well as being designed, built and selected such that it is appropriate for the process operating conditions. 


  1. Processing Magazine.

Non-Newtonian Fluids

Non-Newtonian fluids are encountered in a wide and varied range of industrial processes including food, mining and minerals, pharmaceuticals, pulp and paper processing and other products in the form of pastes, slurries, concentrated solid suspensions or emulsions. Examples include coal slurries, mud, sludge, paint, ketchup, mayonnaise, blood, shaving cream, toothpaste, dilute solutions etc. These fluids regularly exhibit non-Newtonian viscous properties and it is important for the designer to be familiar with the flow behaviour of such fluids, how to characterise the fluid physical properties of these fluids and how to use these properties to predict flow behaviour in process piping systems.  

Newtonian fluids exhibit a linear relationship between shear rate and shear stress. The slope μ (viscosity) is constant and therefore, the viscosity of a Newtonian fluids is independent of shear rate. Non-Newtonian fluids on the other hand typically exhibit either shear thinning or shear thickening behaviour and in some cases, exhibit a yield stress, i.e. a stress level which must be overcome before the fluid begins to flow. There are a number of common rheology models or methods often used to characterise Non-Newtonian fluids. These include the Herschel-Bulkley, Power Law, Casson and Bingham Plastic models, all of which are available in FluidFlow software. 

The viscosity of a Non-Newtonian fluid can take various forms, some of which are outlined in Figure 1.  The flow behaviour classes outline in Figure 1 are Bingham Plastic, Dilatant, Pseudoplastic and Yield Pseudoplastic. 

Figure 1: Non-Newtonian Fluids (Shear Stress vs Shear Rate).

A Bingham plastic fluid exhibits a yield stress which essentially means it behaves as a fluid above the yield stress and like a solid below this stress level. Some non-Newtonian suspensions exhibit these properties such as toothpaste, paint, mayonnaise and shaving cream. 

Dilatant fluids exhibit shear thickening properties, i.e. the fluid viscosity increases with the rate of applied shear stress. A mixture of cornstarch and water is a good reference example of a dilatant fluid.   

Pseudoplastic fluids such as paint and ketchup exhibit shear thinning properties where the viscosity decreases with the rate of applied shear stress. 

Yield Pseudoplastic fluids are similar to Pseudoplastic fluids with the exception that they exhibit a yield stress level, i.e. the fluid behaves as a Pseudoplastic fluid above the yield stress level. 

Figure 2 provides an illustration of typical shear rate vs shear stress data for a non-Newtonian fluid. We can see from the linear graph plot on the left that curve relationship for fluid is y = 0.0197x + 52.68 where the line has a slope of 0.0197 (coefficient of rigidity) and an intercept of 52.67 Pa (Yield Stress). This fluid data has been defined as a Bingham Plastic fluid in the FluidFlow database as shown on the right of Figure 2. We can see the data-points together with the curve equation is shown in FluidFlow providing a relationship of shear stress = 0.0197(shear rate) + 52.68. 

Figure 2: Non-Newtonian Fluids (Shear Stress vs Shear Rate).

FluidFlow helps you characterise non-Newtonian fluids as you can define your data-points of shear rate vs shear stress and swap through the four available rheology relationships (Herschel-Bulkley, Power Law, Casson and Bingham Plastic) and choose the relationship/curve-fit which best represents the defined data-points. A simple yet very effective tool. 

Figure 3 provides an illustration of a non-Newtonian pipe flow system from a mining application as modeled in FluidFlow. The system includes two pumps in parallel transporting a Bingham Plastic fluid through over 8,900 m of pipework with changing elevation. 

Figure 3: Overland Non-Newtonian Pipe Flow System.

Figure 4 provides an illustration of a red mud pipe flow system as modeled in FluidFlow. The system includes eight pumps transporting a non-Newtonian fluid through over 7,300 m of pipework. 

Figure 4: Red Mud Pipe Flow System.

Some typical applications where FluidFlow is used to model Non-Newtonian flow systems include:

  • Mine Tailings Systems. 
  • Metal Concentrate Processing. 
  • Food Industry Applications. 
  • Dredging and Waste Systems. 
  • Medical Devices/Instrumentation (Blood Flow)

The importance of accurate slurry rheology and slurry viscosity data in piping transportation system design.

Fly Ash Viscosity

The characteristics of slurry viscosity are defined by its rheology which describes the deformation and flow behaviour of materials, both fluids and solids. It applies to fluids which exhibit complex microstructures such as sludge and suspensions, mud, fluid-foods etc. These fluid types are regularly encountered in mining and mineral processing industries as well as food processing and pharmaceutical processes.  In order to design and develop accurate and efficient piping systems, a good understanding of the rheology of slurries and slurry viscosity is essential, and determination of the slurry rheology is of paramount importance. 

Slurry rheology is affected by various attributes such as mass fraction, shape, size, density etc of the solid particles suspended in the slurry as well as the density and viscosity of the carrier fluid, typically water. 

It probably comes as no surprise that one of the most important rheological attributes to be considered is the slurry viscosity. The viscosity of a Newtonian fluid, unlike a non-Newtonian fluid, is independent of shear rate. The more common rheology models used to characterise non-Newtonian fluids include the Herschel-Bulkley, Power Law, Casson and Bingham Plastic models, all of which are available in FluidFlow software. 

FluidFlow allows you to plot the curve relationship of shear rate vs shear stress or define the slurry viscosity equation constants. The method you choose is largely dependent on the available slurry rheology data. Figure 1 gives an overview of fly ash defined as a non-Newtonian fluid. In this instance, the Power Law viscosity relationship has been applied effectively to the slurry shear rate and shear stress characteristics. 

Fly Ash Viscosity

Figure 1: Fly Ash Shear Rate vs Shear Stress – Power Law Fluid.

The slurry density is also an important characteristic. The density of a slurry is affected by the density of the carrier fluid as well as the density of solid particles and concentration of solid particles in the slurry. Although the concentration of solid particles is regularly defined by weight percent (Cw), the concentration can be defined as either weight or volume percent in FluidFlow. 

When designing or modelling a slurry flow system, sourcing good slurry rheology data is vital. Having good data combined with a good software tool will help enormously with the sizing and selection of piping, pumps valves and fittings. It will also help greatly with the economic considerations of the project. Insufficient or poor rheology data on the other hand can be a contributing factor in the failure or success of a project. FluidFlow software helps you characterise your slurry based on the rheology and slurry viscosity data you have available. This saves time and helps validate the slurry data giving confidence in the developed system design/model. 

If reliable on-site data, literature data or prior experience is unavailable, the first step is to complete laboratory testing to determine the rheology of the slurry. 

Once reliable data has been sourced, the next step of course is to design the slurry system. Here are some general and useful tips to help with this challenge: 

  • Minimise pipe length to minimise pressure losses.
  • Ensure pipes are as straight as possible to minimise pressure losses and wear.
  • Install pipes in a sloped manner to facilitate emptying of pipes.
  • Pipelines with low velocities should be eliminated in an effort to avoid settling of particles and prevent sanding and ultimately, pipe blockage.
  • Minimise number of pipe fittings. 
  • Install drain valves and flushing inlets. 
  • Pockets and blind areas should be avoided. Where this is not possible, provide a drain valve. 
  • Install wider bends to reduce pressure losses and wear.

This blog discussion has been kept brief in an attempt to target some key points (slurry viscosity and slurry density) in slurry flow design. Designing slurry piping systems can often be a complex challenge, however, the above should help with some of the basics. 

Factors you should consider when sizing plant pipework.

Pipework: One of the most important tasks in plant design is the sizing of the process pipework.

Sizing your pipework incorrectly will have a major impact on plant capital cost, operating costs and plant reliability. 

Piping always represents a significant part of the total installed cost on any plant, typically 10–20% in process plant and up to 80% of the capital cost in long pipeline transport projects. Plant maintenance and energy usage associated with fluid transport add to these costs. As a designer, it is your responsibility, to make the most effective design possible.

With this in mind, we ideally need a way of comparing the economic performance of piping in order to make informed decisions. Pipe sizing decisions should be made by comparing capital and operating costs, which necessitates access to up to date cost data, pipe run lengths and plant design requirements such as flows, temperatures and pressures. Paradoxically in many design projects it is often uneconomic to carry out an economic design of each pipe unless you use the assistance of a suitable software package.

The chart below shows such an approach. Installed costs include material, installation and labor which increase with pipe size. Operating costs, either annualized or summed over the project lifetime, include energy and plant maintenance and decrease with pipe size. By adding the two costs we get a minimum exact economic pipe size – see Figure 1.

Sizing Plant Pipework

There are many computer software applications that help size pipes however, you need to examine the basis for such programs otherwise significant errors will nullify the validity of the program for this purpose. For example, the fluid phase-state and flow characteristics are most important considerations and as such, would you use the same criteria to size a pipe transporting cold water as a pipe transporting steam condensate?

The most beneficial approach is to size the pipework together with other fluid equipment items such as pumps, control valves, heat exchangers etc and in doing so, achieve a total system design. This is feasible only by using specialised software. 

As we have stated, the most complete method for pipe sizing is based on economic velocity but it is worth noting that this approach should be used with caution. There are clearly many situations where pipe sizing to economic criteria should not be used. Some examples are; pump suction pipework, fluids containing solids that may settle, two-phase flow, non-Newtonian liquids, steam condensate lines and pipes that are operated intermittently, an example of such plant would be a fire protection system.

We should also remember that it is not necessary for pipe sizing to be exact. This is because we will usually select a pipe size based on standard pipe sizes. Since the application of pipe sizing rules can never be exact, engineers often use simpler approaches such as sizing to a given pipe velocity or a specific pressure gradient. Fluid flow literature contains numerous references to pipe sizing velocities and/or pressure gradients. Some of the disadvantages of using the simpler methods are that recommended velocities vary with each fluid and each piping material, making the process of always using the appropriate recommended velocity error-prone.

Another disadvantage is that recommended velocities change with time. As an illustration, in the 1980’s the most economic velocity for pumping water through steel pipes was around 2 m/s, 30 years later the economic velocity for pumping water is around 1.3 m/s. This reflects the relative large increase in energy costs compared to capital costs over this time period.

Nevertheless, it is important to select the “best” standard pipe size, so let’s follow the logical steps required to achieve this goal for a Newtonian fluid below its boiling point.

Literature References:

1. Updating the rules for pipe sizing – Durand et al – Chemical Engineering Jan 2010


2. Rules of Thumb for Chemical Engineers 4th Ed – Branan, p7-9.

3. Piping Calculations Manual – Menon.

Common Pitfalls of Large-Diameter Hydraulic Piping System Design

Piping System Design; Much has been written about common problems with poorly designed pumped piping systems. Whilst there is no “one size fits all” or “rule of thumb” solution for all systems types, i.e. liquid, gas, slurry, two-phase etc, the following should serve as a useful baseline consideration for water transmission pipelines.

It can be common practice for engineers to use template design calculation sheets for liquid transmission lines which utilize rules-of-thumb as part of the template for design. This can often result in the failure to acknowledge inherent differences from one project to the next. In large-scale water supply projects, errors in hydraulic system design such as underestimating friction losses can be magnified resulting in reduced system capacity, catastrophic failures or even the potential for litigation.

A common mistake in the design of large-diameter pipelines is underestimating pipe friction losses. A widely used method for determining friction losses is the Hazen-Williams (H-W) equation as it is relatively simple and easy to use. This equation is empirical and in the design of large-diameter piping system, has a limited range of applicability. On the other hand, the Darcy-Weisbach equation provides a better approximation of friction loss as it considers the pipe roughness and Reynolds Number for different pipe materials and is valid across all pipe sizes and turbulent flow ranges. Although there is an abundance of evidence on the limitations of the Hazen-Williams equation, the evidence suggests it is continually misused in industry. 

The Hazen-Williams equation relates the flow of water in a pipe with the physical properties of the pipe and the pressure drop due to friction. The equation uses a pipe roughness coefficient, “Cfactor”, which is not a function of Reynolds number making the H-W equation much easier to use in comparison to other equations. However, the equation can only be used for water, within the transition flow region and near room temperature. The pipe roughness coefficient varies with pipe size and according to published literature, errors can be significant (up to 40%) for pipes less than 8 inches (200 mm) and larger than 60 inches (1500mm) for cold or hot water and for unusually low or high velocities. Notwithstanding this however, the H-W approach has been verified in the field for “common” pipe sizes and conventional flow rates.

The Hazen-Williams equation has been used for over a century and, despite its range limitations there is a large database of available information about the inner surface roughness of pipes as they age.’ Evidence suggests that many hydraulic engineers use the Hazen-Williams formula outside its valid range but researchers have “introduced C-value as functions of pipe diameter and equivalent sand-grain roughness” to make the formula produce reasonable results. A word of caution however for design engineers. It is good design practice to consider a range of Hazen-Williams C-factors when calculating friction losses over the life expectancy of a given pipeline.

The Darcy-Weisbach equation provides a much better approximation of friction loss by comparison with the Hazen-Williams equation as it considers Reynolds number and pipe roughness for different pipe materials and is valid across all pipe sizes and turbulent flow ranges. The equation is dimensionally homogeneous, can be used for any pipe sizes with fluid at any temperature under any flow range and with any liquid/incompressible fluid.

It is prudent when designing piping systems to accurately and conservatively estimate pipeline friction losses however, the engineer needs to also be wary of being too conservative and overestimating friction loss. If the actual operating friction loss is less than estimated at design, the resultant discharge flows will be higher than anticipated. Overestimating friction loss can lead to less efficient operation of pumps which operate in less efficient ranges of the pump performance curves, and installation of larger pumps, motors, pipe, and equipment than what is actually needed for the required flow rates. It is also documented in published literature that a fear of insufficient capacity for future flow rates often leads to the most frequent and worst blunder of all……choosing pumps and pipes too large for the actual flows to be encountered. This can ultimately lead to premature pump failure, poor operating conditions, wasteful energy usage, higher operating costs and higher capital installation costs.

Poor design on multi-million dollar water transmission systems can be disastrous to system operators and owners and cost million of dollars extra due to reduced water supply, increased water rates and required public expenditure on additional water supply sources. It is essential that a designer is aware of the inherent differences between each project and avoid applying a rule of thumb approach or one fits all spreadsheet to calculate the system conditions.


1. ASCE Pipelines 2011.

Common Pitfalls of Large-Diameter Hydraulic Piping System Design

Common pitfalls of large-diamter hydraulic piping system design. Much has been written about common problems with poorly designed pumped piping systems. Whilst there is no “one size fits all” or “rule of thumb” solution for all systems types, i.e. liquid, gas, slurry, two-phase etc, the following should serve as a useful baseline consideration for water transmission pipelines. 

It can be common practice for engineers to use template design calculation sheets for liquid transmission lines which utilize rules-of-thumb as part of the template for design. This can often result in the failure to acknowledge inherent differences from one project to the next. In large-scale water supply projects, errors in hydraulic system design such as underestimating friction losses can be magnified resulting in reduced system capacity, catastrophic failures or even the potential for litigation. 

A common mistake in the design of large-diameter pipelines is underestimating pipe friction losses. A widely used method for determining friction losses is the Hazen-Williams (H-W) equation as it is relatively simple and easy to use. This equation is empirical and in the design of large-diameter piping system, has a limited range of applicability. On the other hand, the Darcy-Weisbach equation provides a better approximation of friction loss as it considers the pipe roughness and Reynolds Number for different pipe materials and is valid across all pipe sizes and turbulent flow ranges. Although there is an abundance of evidence on the limitations of the Hazen-Williams equation, the evidence suggests it is continually misused in industry. 

 The Hazen-Williams equation relates the flow of water in a pipe with the physical properties of the pipe and the pressure drop due to friction. The equation uses a pipe roughness coefficient, “C-factor”, which is not a function of Reynolds number making the H-W equation much easier to use in comparison to other equations. However, the equation can only be used for water, within the transition flow region and near room temperature. The pipe roughness coefficient varies with pipe size and according to published literature, errors can be significant (up to 40%) for pipes less than 8 inches (200 mm) and larger than 60 inches (1500mm) for cold or hot water and for unusually low or high velocities. Notwithstanding this however, the H-W approach has been verified in the field for “common” pipe sizes and conventional flow rates. 

The Hazen-Williams equation has been used for over a century and, despite its range limitations there is a large database of available information about the inner surface roughness of pipes as they age.’ Evidence suggests that many hydraulic engineers use the Hazen-Williams formula outside its valid range but researchers have “introduced C-value as functions of pipe diameter and equivalent sand-grain roughness” to make the formula produce reasonable results. A word of caution however for design engineers. It is good design practice to consider a range of Hazen-Williams C-factors when calculating friction losses over the life expectancy of a given pipeline. 

The Darcy-Weisbach equation provides a much better approximation of friction loss by comparison with the Hazen-Williams equation as it considers Reynolds number and pipe roughness for different pipe materials and is valid across all pipe sizes and turbulent flow ranges. The equation is dimensionally homogeneous, can be used for any pipe sizes with fluid at any temperature under any flow range and with any liquid/incompressible fluid. 

It is prudent when designing piping systems to accurately and conservatively estimate pipeline friction losses however, the engineer needs to also be wary of being too conservative and overestimating friction loss. If the actual operating friction loss is less than estimated at design, the resultant discharge flows will be higher than anticipated. Overestimating friction loss can lead to less efficient operation of pumps which operate in less efficient ranges of the pump performance curves, and installation of larger pumps, motors, pipe, and equipment than what is actually needed for the required flow rates. It is also documented in published literature that a fear of insufficient capacity for future flow rates often leads to the most frequent and worst blunder of all……choosing pumps and pipes too large for the actual flows to be encountered. This can ultimately lead to premature pump failure, poor operating conditions, wasteful energy usage, higher operating costs and higher capital installation costs. 

Poor design on multi-million dollar water transmission systems can be disastrous to system operators and owners and cost million of dollars extra due to reduced water supply, increased water rates and required public expenditure on additional water supply sources. It is essential that a designer is aware of the inherent differences between each project and avoid applying a rule of thumb approach or one fits all spreadsheet to calculate the system conditions. 


1. ASCE Pipelines 2011. 

Slurry Piping Systems in Mining & Mineral Processing Industries

Slurry Piping Systems: The mining and mineral processing industry provides a diverse set of challenges in the design and operation of pipelines, from complex slurry pipelines through to mine dewatering systems. The traditional requirements of reliability and availability remain key drivers in the design of any system. However, the need to address capital cost, efficiency and optimisation issues continues to become more prevalent in the face of rising production costs and environmental responsibilities. Increasingly, the success of piping systems design is being assessed against all these criteria and a good designer will have considered these aspects during the design process. 

FluidFlow’s powerful capabilities have been used successfully by mining and mineral processing companies for years to design complex fluid systems, allow design teams to achieve a consistently high level of accuracy and a thorough understanding of system behaviour throughout the entire design development stages.

FluidFlow in particular addresses the needs of the mining and mineral processing industry which has resulted in a strong presence globally. 

Considering slurry flow systems specifically, FluidFlow helps engineers characterize slurries (settling or non-settling/non-Newtonian), evaluate pump performance, consider pump de-rating, predict particle settling velocities, predict likelihood of blockage, predict flows and pressures, estimate solids delivered in the slurry etc. The software equips the engineer with all the tools required to successfully develop efficient slurry flow systems. 

Typical piping systems encountered in the mining industry include tailings and paste systems. Tailings is a bi-product of the processing of ore. When the ore is processed, the valuable minerals are separated leaving the fine-grained waste material known as tailings. The tailings are usually pumped from the thickeners through a slurry transportation pipeline and deposited at a dedicated tailings management/storage facility. 

The design of tailings piping systems can be quite complex as the properties of the tailings such as concentration etc can fluctuate. It is therefore often necessary to design and develop a piping transportation system which can accommodate the anticipated variations in slurry characteristics. This is where FluidFlow has helped designers understand and overcome these operational issues. Figure 1 provides an illustration of typical tailings piping system as developed in FluidFlow.  

Figure 1: Tailings Systems.

Another piping system regularly encountered is mine dewatering. The control of ground water and dewatering of underground or open pit mines can be a very important aspect mine operation. 

Figure 2 provides an illustration of typical mine dewatering system as developed in FluidFlow. This system shows six pumps on a pontoon providing the initial lift to the upper elevation where the water is then transported via 17 km of overland piping to the final location. 

Figure 2: Mine Pit Dewatering.

Dust extraction systems can also be designed in FluidFlow. The model shown in Figure 3 was used to size the ducting and fan required to remove dust from a conveyor system taking ROM product from a primary underground crusher to a secondary crushing circuit hopper. The conveyors ran in restricted tunnels and had hoods located over the centreline of the conveyors to remove dust at various points in the system.

Figure 3: Dust Extraction System.

Some typical applications where FluidFlow is used in the mining and mineral processing environment include:

  • Flotation circuits (including launder flow and froth factor de-rating).
  • Grinding circuits.
  • Thickening circuits.
  • Process water circuits.
  • Compressed air circuits (instrument and process).
  • Dust suppression and extraction systems.
  • Pressure and Heap leach spray systems.
  • Magnetite recovery systems.
  • Dewatering circuits.
  • Heat exchange circuits.

This blog broadly discusses just a few sample piping systems in the mining and mineral processing industries. 

FluidFlow is trusted among many of our users in countries such as Canada, Africa, Madagascar and Australia to mention a few. 

Operating Efficiency of Centrifugal Pumps

We are all fairly familiar with the performance curves for centrifugal pumps but what may not be so obvious is the extremely important relationship between the actual duty-point on the pump curve and the resultant pump reliability. Overlooking this relationship can result in reduced equipment lifespan, higher maintenance costs and quite costly pump repairs not to mention the headaches and associated costs that can arise from unexpected and preventable equipment downtime. 

Centrifugal pumps should be selected to operate as close to the best efficiency point (BEP) as possible. But what is the BEP? The best efficient point represents the duty point on the pump performance curves where the highest level of operating efficiency is achieved for a given impeller diameter.  In an ideal scenario, a centrifugal pump would operate at or as close to the BEP as possible at all times however, fluctuating demand profiles and system upsets are common causes for the actual duty point to move to/from the BEP. 

Let’s consider the performance curves for a sample centrifugal pump as shown in Figure 1. The BEP is denoted by the green duty-point, i.e. the apex of the efficiency curve. Therefore, if this specific pump was operating in a system whereby the duty-point on the pump capacity curve was 16 m3/h with a head requirement of 68 m fluid, the pump would be operating at the BEP (71 %) and therefore exhibiting the highest level of operating efficiency for the given impeller diameter on which these performance curves are based, in this case 160mm. 

Figure 1: Centrifugal Pump Performance Curves.

Let’s consider the operating condition in this example of 16 m3/h @ 68 m fluid as our ideal duty-point or operating condition. If the flow rate in the system fluctuates, then so too will the duty point on the performance curves. For instance, operating at a higher flow rate than the BEP will result in the duty point on the performance curves moving to the right of the BEP. Likewise, operating at a lower flow rate than the BEP will result in the duty point on the performance curves moving to the left of the BEP. The key point in both scenarios is that the duty point moves away from the BEP which of course is away from the apex of the efficiency curve and as such, reduces the overall operating efficiency of the pump. This will result in increased energy costs which are sometimes considered “insignificant” unless the pump operates under these conditions for extended time-periods or has a large horsepower motor.  

In real-world scenarios, it is generally not possible to operate exactly at the BEP at all times. However, if the pump can be sized and selected to operate within an optimum operating range of the BEP, the efficiency and pump reliability can be maximised. Some literature sources in the public domain quote a range of 70 to 120 % or 80 to 120 % of the BEP.  

It is generally considered that a centrifugal pump should be sized and selected to operate within the range of +/-10% of the BEP. This should serve as a useful guide for pump selection. 

Figure 2 outlines the ideal operating range for this specific sample pump based on +/- 10% of the BEP (14.4 to 17.6 m3/h).

Figure 2: Ideal Operating Range of Centrifugal Pumps.

So, we have discussed the general effects of operating away from the BEP but what are the ultimate effects on the pump as we operate further away from the BEP operating range. Figure 3 gives an overview of the conditions which arise when a centrifugal pump operates further away from the BEP for an extended period of time. This is true for incorrectly sized or poorly matched pumps. 

Figure 3: Ideal Operating Range of Centrifugal Pumps

Figure 3 demonstrates that the pump will encounter cavitation, elevated temperatures, lower bearing and seal life etc. The pump will also experience excessive vibration, potential lubricant breakdown, accelerated wear on components such as shafts, bearings and impellers. 

Note, suitable software tools can help with the correct sizing and selection of pumps. Figure 4 shows the performance curves for a vendor pump on the left-hand side. This pump data was then modeled in a sample test-case system in FluidFlow. When using FluidFlow, we can see where the system curve intersects the pump capacity curve and the actual duty point on the performance curves. The calculated results for the pump allow us to see the operating efficiency, NPSHa, NPSHr as well as the duty head requirement. This data helps us select a pump best suited to the system. Changes in pump speed and impeller diameter can also be considered as well as the effect of fluctuating demand on the system. 

Figure 4: Modeling Centrifugal Pumps in FluidFlow.

FluidFlow includes powerful pump and equipment sizing functionality which helps simplify the design process of sizing pipes, control valves, relief valves, orifice plates, pumps etc. Perhaps more importantly, this feature is proven to drastically reduce required design time on projects meaning designers can use their time more effectively as well as have confidence in the pumps and equipment sized and selected for the system under design. 

A Study of Choked Flow in Gas Piping Systems

The phenomenon of choked flow is often encountered in gas piping systems and tends to occur where there is a change in the flow path cross-sectional area. It can, therefore, occur in locations where we have orifice plates, valves, and fittings. It can however, also occur as flow exits a pipe into a vessel or to atmosphere.

Sonic choking is experienced in pipelines where the flowing gas reaches a velocity which is equal to the local sonic velocity of the gas at that given temperature and pressure. The gas, therefore, attempts to accelerate beyond the local sonic velocity and therefore becomes limited or choked.

If we consider air flow systems, choking typically occurs when the downstream absolute pressure (P2) is 52.8% of the upstream absolute pressure (P1), i.e. the absolute pressure ratio is 0.528.

Let’s initially consider a scenario where we have air flowing through a 10M long 2-inch pipeline with an inlet and outlet pressure of 6 & 5 bar a respectively. The air in the pipe stream must expand due to the pressure loss in the line. This results in a reduction in air density which tends to increase the fluid velocity and as such, the actual volumetric flow rate in the system. See Figure 1 below for details.

Screen Shot 2018-06-26 at 14.29.07

Figure 1: Air Flow System.

If we gradually decrease the downstream pressure we can reach the choke point for the system. Based on our rule of thumb, this should occur when the outlet pressure is in the region of approximately 3.168 bar a.  Let’s recalculate the system based on an arbitrary outlet pressure of 2.0 bar a.

In Figure 2, we can see the result of this design scenario. FluidFlow has automatically detected endpoint choking conditions in the pipeline.

Screen Shot 2018-06-26 at 14.48.25

Figure 2: Choked Flow with Outlet Pressure 2.0 bar a.

We can see from Figure 2 that the Mach number is 1.0 and the mass flow rate is 1.7146kg/s. Let’s see what happens when we reduce the outlet pressure even further to 1.0 bar a.

Screen Shot 2018-06-26 at 14.40.23Figure 3: Choked Flow with Outlet Pressure 1.0 bar a.

As we can see, the mass flow is unchanged as it has become choked in this system.

FluidFlow automatically detects choked flow and also plots the resistance, density, static pressure, static temperature & velocity curves for all pipes. Figure 4 provides an illustration of the density curve for the pipe in Figure 3.

Screen Shot 2018-06-26 at 14.33.56

Figure 4: Density vs Pipe Length Curve.

Figure 5 illustrates how FluidFlow detects choking conditions across a size-change component such as an orifice plate. In this case, we have an inlet and outlet pressure in a 10.0M 2 inch pipeline of 2.2 and 1 atm respectively. The orifice diameter is only 10mm. The pressure loss across the orifice plate is quite high at 1.2 bar. As a result, the air must accelerate to satisfy the continuity equations. In this case, the gas attempts to accelerate beyond the local sonic velocity causing choked flow.

Screen Shot 2018-06-26 at 14.35.16

Screen Shot 2018-06-26 at 14.42.34

Why choose the FluidFlow Gas Module ?

FluidFlow does not assume gas ideality but solves for real gas conditions. The software will solve complex systems including gas mixtures (by mole or mass %). FluidFlow is provided with heat transfer functionality and a comprehensive database of fluids and components as standard.

Designing Compressed Air Systems

Inadequate or poorly designed compressed air distribution systems can lead to low productivity, poor air tool performance and perhaps more importantly, high energy bills. In order for a compressed air system to operate properly and cost effectively, it should be carefully designed to meet the needs of your applications. As part of the design process, there are six items which should be considered and factored into the final system design to give optimum results at maximum efficiency. The six items include demand, compressed air quality, supply, storage, distribution and control or management and all six must work together for the system to achieve top performance levels. The following discussion considers a number of these items.

1) Compressed Air Demand
To determine the demand for new systems, the operating pressure requirements and duty cycle of individual equipment should be considered. Compressed air consumers are rated by the manufacturer for optimum performance at a certain pressure and air flow rate. To design a system that delivers uniform pressure, it is necessary to ensure all tools and equipment work efficiently within a narrow pressure range. However, if this is not achievable, the system can be designed to operate at a higher pressure and user regulators to reduce the pressure as required. A booster can also be added to increase pressure for any specific applications requiring higher pressures. One final option available it to design two separate compressed air systems operating at different pressures. Leakage and artificial demand can often represent a significant amount of the system overall demand. All compressed air systems exhibit leaks and leakage can be measured in a number of ways while no pneumatic equipment is in use; measured using the loaded running time of a compressor, timing the pressure drop of the receive tank while all compressors are off or
measuring leakage at the point of use. Compressed air system demand can also include artificial demand caused by excess system pressure that doesn’t increase productivity. Artificial demand can be reduced significantly by installing a regulator at the point of use of a flow controller at the beginning of the distribution network.

2) Compressed Air Quality
The quality of compressed air is determined by measuring three main contaminants which include water vapor content as measured by pressure dew point temperature, oil content as measured by concentration and finally, solid particles as measured by the concentration of by their size. The level of contamination is influenced by the type of compressor, dryer, filtration and other related components. It’s no surprise then that the higher the air quality required, the more expensive the equipment. The International Standard document ISO-8573-1 provides a classification system for the main contaminants of a compressed air system and identifies how other contaminants can be identified in addition to the classification system.

Compressed air for power-tool usage, sand blasting, pneumatic pumps etc is a relatively low-grade quality where water, oil or solid particles in the air are more of an annoyance than a major concern.

Compressed air for instrumentation is a higher quality which is used in more sensitive areas where water or particulates could contribute to significant quality issues in the process. Paint spraying and powder coating are good examples of such a process where contaminants will ultimately affect final product quality. This classification of compressed air would be filtered for solids particles and oil and dried to a higher standard to that of compressed air for say, power-tools.

Process air is often used in food or drug production. Compressed air for use under these circumstances would need to be completely oil-free with almost negligible water vapor.

Higher quality compressed air is used in hospitals or for diving applications where the air must be of a quality suitable for safe breathing.

Compressed Air Supply

The compressed air supply must match the compressed air demand. If the supply, storage, or distribution system are not optimized. Excessive pressure fluctuations can occur resulting in increased operating costs and reduced productivity.

Many compressors are controlled by the line pressure. An increase in demand results in a drop in line pressure which is then rectified by an increase in the output of the compressor(s). A rise in line pressure, therefore, indicates a reduction in downstream demand which causes a reduction in compressor output. There are various forms of compressor control which can be utilized to manage such system operating conditions.

Single compressor installations are generally more suited to smaller compressed air installation where the compressor may often run at full capacity. Multiple compressors installations can offer many advantages such as; an ability to adjust to changing usage patterns, provide flexibility, floor space flexibility, can be centralised or de-centralised and of course, by their nature offer a backup facility in the event of plant failure.

4) Compressed Air Storage

The provision of sufficient storage is vital and represents available energy that can be utilized and replenished as required and at any time. The air receiver tank generally provides the bulk of the total storage capacity. In some cases, the compressor controls depend on storage to limit maximum cycling frequency when the demand is below 100% of supply. A correctly sized tank will prevent excessive cycling. 

Correctly sized receiver tanks can also provide sufficient storage capacity for any peaks in demand. During peaks demand periods, a poorly designed system can experience a drop in pressure as air in excess of system capacity is drawn from the system. As not all compressors in a multi-compressor system remain online at all times, the actual air supply at any time can be less than the total system capacity. During the time required to bring additional compressors online, the stored compressed air can be used to prevent any pressure drop in the system. The quantity of stored capacity needed is dependent on the amount of excess demand in cubic feet, available pressure differential between the compressor station and demand point, compressor start-up time as well as the time available to replenish the stored compressed air.

The installation of a flow controller downstream of the receiver tank is essential for providing additional compressed air when needed without downstream pressure fluctuations. The flow controller works life a precision regulator, increasing or reducing flow to maintain constant line pressure. It also provides the necessary pressure differential between the receiver tank and the system to create storage without changing system pressure downstream.

5) Compressed Air Distribution

An ideal distribution system provides a sufficient supply of compressed air to all demand points at the required pressure. Inadequate or poorly designed compressed air distribution systems can lead to low productivity, poor equipment performance and high energy bills. When designing a compressed air system, it is therefore good practice to consider the factors which help improve the efficiency and reliability of the compressors and ancillary equipment, minimise leakage and pressure drops and improve compressor life-cycle cost.

In general, there are three demands imposed on a compressed air distribution system; low-pressure drop between the compressor and most remote demand point, minimum leakage from the distribution pipework and efficient condensate separation if a compressed air dryer is not included in the system.

The compressed air pipe routing, design and dimensions are important factors for the efficiency, reliability and cost of compressed air production. Occasionally, a large pressure drop in the piping system is compensated by adjusting or increasing the working pressure of the compressor from for example, 7 bar to 8 bar. This results in inferior compressed air economy. Furthermore, when compressed air consumption is reduced, the pressure drop is also reduced and so the pressure at the demand point consequently rises above the allowed level. This therefore is not a recommended operating strategy.

In general, fixed compressed air distribution systems should be sized such that the pressure drop in the pipes does not exceed 0.1 bar (10,000 Pascals or 1.45 psi) between the compressor and most remote demand point. The pressure drop arising from flexible hoses, couplings and fittings should also be included in the pressure drop calculation. However, if possible, in an attempt to reduce pressure losses, the number of bends, valves, fittings or flow obstructions should be kept to a minimum.

A useful approach is to design a ring main or looped piping system to serve the space where the compressed air consumption will take place. This is an effective way to minimise pressure drop in a system. Branch pipe connections are then run from the loop to serve the various demand points. This helps provide for a uniform compressed air supply as the air distributed to the demand point from two directions.

The design of a ring main system is a recommended approach however, may not be entirely suitable in scenarios where there are large compressed air consumers located at a much greater distance from the compressor installation. It is recommended that a separate compressed air main should be routed under these circumstances to serve this equipment.

Compressed air piping system can generally be grouped into four main parts; risers, distribution pipes, service pipes and compressed air fittings. The risers transport the compressed air from the compressors to the consumption zones, the distribution pies split the air across the distribution zone and the service pipes route the air from the distribution pipes to the working areas/final demand points.

At design stage, the starting point is generally to obtain a diagram indicating the location of all compressed air equipment and an equipment list detailing all compressed air consumers. The compressor plant should be in a centralized location in close proximity to all relevant applications and processes in order to minimize the length of piping between the compressor station and the consumers. The consumers should ideally be grouped in a logical manner and supplied from the same pipe.

The location of the compressor must also give consideration to the quality of intake air, ideally clean, dry and cool. The compressors should be positioned clear of any potential steam, chemical vapors, engine exhaust and dust.

The pressure developed by the compressor plant can generally never by fully utilized as a result of pressure losses occurring in the system from pipework and fittings. When designing and sizing the various piping zones of the system, the following rule of thumb can be used for pipe losses.


Screen Shot 2018-06-19 at 12.53.04

The length of the pipes required for the system can be scaled off the design plan and section drawings for the building/facility. The pipes can then be sized using relevant formulae and/or pipe sizing tables or nomograms. The flow rate, pressure, allowed pressure drop and length of pipe usually must be known in order to correctly size the pipework. When sizing the pipework, the equivalent length for each pipe section is estimated in an attempt to take into account the losses from fittings.  It is also often necessary to design and calculated the service pipes, distribution pipes and riser pipes separately, particularly for larger installations.

It is also worth remembering that, horizontal pipe installations should have a slope of around 1 to 2 % toward the point of consumption to aid the transfer of condensate to the predetermined drain points. Note, some would argue that the need to slope the pipework is no longer required when a correctly sized air dryer is installed however, the cost is minimal/negligible and offers added protection in the event of an air dryer failure.

The use of an appropriate software tool helps overcome the difficulties with designing these systems by hand. Using tables, nomograms etc can be time-consuming and occasionally, error prone. Applying the equivalent length approach in a hand calculation also produces a lower level of accuracy by comparison to a software-based solution. FluidFlow helps with the design of such systems and allows the designer to consider dynamic type operating scenarios with fluctuating demand etc.


Compressed Air FF

Figure 1: Compressed Air System Model.

Compressed Air System Performance

Tests and experiments have shown that for every 1 bar in pressure drop across the system, the resulting power consumption is increased by circa 6 – 7 % (plus the additional cost of unregulated users). Accordingly, the distribution network can have a profound influence on the system performance. There are a number of steps which can be taken to optimise the air distribution system.

1. Pressure Drops and Line Improvements.

As outlined previously, a ring main is generally the most efficient type of distribution layout. Air mains are usually sized on velocity and a velocity level of 6 to 9 m/s is common as this is sufficiently low to prevent excessive pressure drop and should also allow reasonable water separation. The local feeding mains can flow up to around 15 m/s. However, in order to prevent adverse pressure drops, the flow velocity in the main header sections should not exceed circa. 6 m/s.

It is best to replace any tee connections for directional angle entry connections or swept tees. Turbulence caused by a 90o tee connection can cause pressure drops resulting in back pressure sending a false “unload signal” to the compressors which can potentially cause excessive cycling of the compressor.

Incorrect pipe sizing and restrictions are a major source of pressure losses in the system. Losses in the interconnecting distribution pipework between the compressor and the header distribution piping are commonplace however, the losses along these lines should be kept to a minimum.

Interconnecting piping between compressors or systems often require close attention. It is vital that they are carefully designed to avoid sending back any false signals to the compressor.

2. Leak Detection

Leaks in compressed air systems are a regular feature. The energy requirements served by compressed air systems are intermittent in nature, however leaks are constant and surprisingly, potentially significant. The monetary cost of leaks can be quite startling and perhaps a little eye watering.

For instance, one 4mm hole in a compressed air distribution pipe can cost €2,005 per annum on a typical compressed air system operating throughout the year and at 8 bar.

In addition to the monetary cost, leaks can cause significant pressure drops resulting in excessive compressor cycling. In an attempt to reduce the pressure loss in a system where excessive leakage is an issue, operators occasionally increase the system discharge pressure. However, this has the effect of exacerbating the problem by increasing the leakage rate and create more leaks in the future.

It is not uncommon for leakage rates to be around 20 – 30%. Leaks can occur at any point in the system with joints, drains, valves regulators etc being the most common sources. Fixing leaks in the most basic form can occasionally be as simple as tightening connections or applying a sealant at a strategic point. However, leaks will be found which require replacement of faulty components. It is worth noting that one of the most effective means of reducing leakage is to reduce the distribution pressure. Note, a 10% reduction in leakage would often be achieved through carrying out an appropriate leak reduction programme. 

3. Dedicated Air Receivers for End Users.

In industrial applications where air pressure is subject to large fluctuations or variations, air receivers are beneficial. In these situations, the increased compressed air requirement is compensated by air from local air receivers thus minimising idling at the generation station.  The receivers subsequently replenish slowly using control valves to minimise peak energy demand on the compressor station. In addition to reduced compressor cycling, air receivers provide protection for end users that require high pressure by minimising the system pressure drop off while supporting the speed of transmission response in supply.

Other useful tips with regard to compressed air system design include:

  • Position one air receiver near the compressor to provide a steady source of control air, additional air cooling and moisture separation. A large storage receiver can be located downstream of the dryer and filters to act as a buffer for demand surges and controlled by a flow controller.
  • Compressors should be positioned in a clean, dry, cool and well-ventilated room. Ensure there is sufficient room around the compressors and equipment for proper air flow. Manufacturer’s will often specify the minimum spacing required around compressors.
  • Piping in a loop is recommended will all piping sloped to accessible drain points. Air outlets should be taken from the top of the main lines to prevent moisture entering the outlet.
  • Under average conditions, every 100 cfm of air compressed to 100 psig (6.89 barg) produces approximately 20 gallons (75.7 litres) of condensate per day which needs to be treated.
  • The minimum amount of storage recommended is one gallon per cfm of capacity. This should be increased to 4 to 10 gallons per cfm of capacity for systems with sharp changes in demand.


  1. Chemical Engineering, March 2018.
  2. SEAI – Compressed Air Technical Guide.
  3. Atlas Copco Compressed Air Manual.
  4. Designing Your Compressed Air System – Kaeser Compressors.

FluidFlow v3.44

General Release summary: This release provides speed improvements at start up and across a network. There are over 250 new components, centrifugal and positive displacement pumps, controllers, fluids and pipes added to the databases. This release also contains bug fixes.

General Release info:


* Hundreds of new database entries. Over 180 centrifugal pump models added from KSB, Warman etc. Over 70 positive displacement pump models from SPXFLOW & Fristam added.

* Steel pipes database has been expanded to include additional B36.10M sizes and classifications for existing and new steel pipe sizes.

* 7 new control valves and new fluids including d-limonene, thionyl chloride, AVCAT, AVGAS, and Duratherm XLT-120

* Additions and corrections to help file.


* Velocity Pressures now shown without gauge or absolute unit reference.


* Supercritical -> Liquid and Supercritical -> Gas phase change via a fixed temperature change at any element sometimes predicts incorrect exit phase and incorrect heat loss for a fixed temperature change.

* Now allowable to have viscosity AND affinity law correction together. Was a bug that prevented the 2 effects from occurring together.

* Tee junctions with a fluid change caused instability and non-convergence. Now fixed.

* Intermittent bug in print preview that sometimes caused an error is fixed.

* Fixed head unit inconsistency in EGL and HGL charts.

* For old files containing Liquid Relief valves, the Kd was sometimes not updated for an autosize calculation.

* Fixed scripting error on start-up when Scripting module is not available.

* Added check to stop file being saved to a non-existent folder.

Joule-Thomson Effect (J-T)

To obtain maximum accuracy in gas calculations, particularly in systems where large pressure losses occur, the Joule-Thomson effect needs to be included in the solution. The J-T effect describes the increase or decrease in temperature of a real gas (as opposed to an ideal gas) or liquid as it flows from an area of high pressure to an area of low pressure without heat transfer to or from the fluid and no external mechanical work extracted from the fluid. An example of such a scenario would be a gas flowing through a perfectly thermally insulated throttling device such as a valve, nozzle or orifice plate.

The temperature change experienced during a Joule-Thomson expansion is quantified by the Joule-Thomson coefficient.

Joule-Thomson Coefficient

The temperature change arising from a reduction in pressure at constant enthalpy in the Joule-Thomson process can be expressed as:

μJT = (∂T/∂P)H

The value of μJT (oC/bar) depends on the gas type as well as its pressure and temperature before expansion. This coefficient may be either positive (cooling) or negative (heating). For all real gases, the value of μJT will equal zero at some point called the inversion point and the J-T inversion temperature is the temperature where the coefficient changes sign (where the coefficient equals zero).

The inversion temperature represents the temperature of a gas at which a reduction in pressure causes no temperature change. Above this temperature, the gas heats on expansion and below this temperature, the gas cools on expansion. When a real gas expands through a throttling device, the gas temperature may either increase or decrease depending on the initial gas pressure and temperature. Figure 1 provides an illustration of a typical temperature inversion curve.

Figure 1

Figure 1: Temperature inversion Curve.

In gas expansion, the gas pressure reduces and thus the sign of ∂P is always negative. The following table explains when the Joule-Thomson effect heats or cools a real gas.

Screen Shot 2018-05-30 at 14.11.04

It is worth noting at this point that, there is no Joule-Thomson effect for an ideal gas. The value of μJT, therefore, equals zero, i.e., in theory, ideal gases neither heat or cool during expansion at constant enthalpy.

FluidFlow takes account of the Joule-Thomson coefficient when solving gas flow systems using real gas conditions. Figure 2 details the calculation of Joule-Thomson coefficient for a pressure control valve in a pipeline transporting a mixture of n-butane and n-pentane at a pressure of 11bar and temperature of 390K.

Figure 2 J-T PRV

Figure 2: FluidFlow J-T Coefficient Calculation.



Pressure Drop in Natural Gas Pipelines

A previous blog post considered the applicability or the ideal gas equations for gas flow modelling and the potential pitfalls of using such a solution approach. This post considers a number of simple natural gas pipe flow systems which have been solved using FluidFlow. FluidFlow software doesn’t make simplifying assumptions and solves for real gas flow conditions using an equation of state.

Figure 1 provides an illustration of a horizontal natural gas pipe line consisting of 50 miles of 26 in OD pipe having an ID of 25.44 in. The temperature of the flowing gas is 60 F. The upstream and downstream pressures are 825 and 565 psia respectively.

horizontal natural gas pipe line

Figure 1: Modeling Horizontal Flow of Natural Gas.

The solution in Figure 1 shows good agreement with published literature (1) where Q was estimated by way of a hand-calculation to be 462 MMft3/day at 14.65 psia (101008 Pa) and 60 F (15.5 oC). FluidFlow has established a flow rate of 458. 7 MMft3/day. The difference in flow can be attributed to the difference in density of the flowing gas and reference volumetric flow conditions, i.e. a 101325 Pa and 15 oC.

Figure 2 provides an overview of a gas-distillate well pressure model where the inlet pressure has been calculated. The vertical pipe length is 5790 ft with a corresponding well depth of 5790 ft. The pipe ID is 1.995 in with an inlet flow rate of 5.153 MMft3/day (STP) and temperature of 83 F.

Figure 2 gas-distillate example page 323

Figure 2: Gas-Distillate Well Pressure Model.

Published literature (1) makes reference to a hand-calculated inlet pressure of 2,556 psia. FluidFlow has established an inlet pressure of 2,526 psia. A plot of the pressure traverse for this vertical system has also been provided (Figure 3).


Figure 3

Figure 3: Pressure Traverse of Flowing Gas Well.

To make a comparison, two further test case scenarios were created in FluidFlow, one with a well depth of 1,000 ft and one at 3,000 ft. Based on the graph plot, the corresponding pressure is approximately 2,195 psia and 2,340 psia. The models for the two cases are outlined in Figure 4.


Figure 4 vertical pipes

Figure 4: Gas-Distillate Well Pressure Model (Well Depths of 1,000 ft & 3,000 ft).

The FluidFlow solution has established calculated inlet pressures of 2193.7 and 2334.7 psia respectively. Again, the solution produces good agreement with published literature.

Figure 5 provides an illustration of a natural gas distribution system as modeled using FluidFlow. The system consists of 77,103 ft of pipework ranging from 2 to 18 inches in diameter. The initial condition is 50 barg and 100 oC and all outlets are set to 140 barg. Four compressor stations have been modeled in the system generating a total flow rate of 2,000,000 kg/h. The analysis of this system included optimisation of the plant and an evaluation of the effect of pipe roughness after many years of use.

Figure 5 natural gas distribution system

Figure 5: Natural Gas Distribution System.


The Calculation of Pressure drop in the flow of natural Gas through a pipe – Fred H. Poettmann.

To read our previous blog post on Compressible Gas Flow In Pipes; please click here –

Compressible Gas Flow in Pipelines

Many equations used for the solution of gas flow in pipelines do not take into consideration rigorously the deviation of natural gas from ideal gas behaviour. For low-pressure gas flow, the error is considered low, however, for high pressures, high flow rates, low temperatures or the closer the gas is to change in phase state, this error can be significant. This deviation from ideality must be taken into consideration. It is therefore important when modeling gas pipe flow systems to use a
software tool which doesn’t make the simplifying assumptions of gas ideality.

FluidFlow does not make these simplifying assumptions but solves for real gas conditions using an equation of state. The simplified application of the ideal gas law (Pv = nRT) makes several assumptions as outlined below;

– Particles in an ideal gas are in constant, random, straight-line motion.
– The volume occupied by the particles is negligible relative to the volume of the enclosure.
– The collisions between particles are elastic with no loss of kinetic energy.
– There are no intermolecular forces acting between the particles in an ideal gas.

An ideal gas assumes a large number of point particles colliding elastically. It neglects any short-range intermolecular forces resulting from repulsion or attraction due to molecular charges and the fact that molecules have a finite volume, i.e. are not infinitely small! This means a real gas is not infinitely compressible whereas an ideal gas has no such limitations.

Although the ideal gas law may be a useful simplified description of gases, all real gases fail to obey this relationship to some degree. The extent to which a real gas departs from ideal behaviour can be seen by rearranging the ideal-gas equation to solve for n (PV/RT = n). This form of the equation tells us that for 1 mol of an ideal gas (n=1), the quantity PV/RT equals 1 at all pressures. However, under the conditions of real gas flow outlined earlier, a product of PV divided by RT is no longer equal to one.

In the graph shown in Figure 1, we can see the deviation of nitrogen gas from ideal gas behaviour. On the y-axis, we have the product of PV/RT. On the x-axis, were have pressure. The dashed blue line shows the behaviour of an ideal gas for which the product of PV/RT is equal to one under any conditions. The orange, grey and yellow dashed lines show the deviation of nitrogen gas from ideal gas behaviour at different temperatures and pressures. These dashed lines show a considerable deviation of ideal gas behaviour. Notice how the conditions that produce this curve are high pressure and low temperature.

Figure 1 nitrogen real gas pv rt graph

Figure 1: Effect of Temperature & Pressure on the Behaviour of Nitrogen Gas.

As high pressures, the deviation from ideal behaviour is considerable and is different for each gas. Real gases, therefore, do not behave ideally at high pressures.

It is noted in many literature sources that the ideal gas equations can be used with some degree of accuracy under certain specific conditions, i.e. at low pressures. Caution should, however, be exercised when applying gas ideality and a thorough check of system operating conditions to ensure the design solution falls within the criteria.

FluidFlow doesn’t make the simplifying and often error-prone assumption of applying gas ideality but solves for real gas conditions using an equation of state. The software, therefore, takes into account the gas compressibility factor (Z) and solves gas flow systems using a marching algorithm for incremental pipe length. This ensures a much higher level of solution accuracy.

The solution of gas flow systems becomes more complex when dealing with gas mixtures and large systems. It is recommended that a suitable software tool should is selected to design and solve gas flow systems in an attempt to eliminate the potential for errors resulting from the simplification of ideal gas flow.

In conclusion, let’s ponder the following quote:

“Perfect gases are like perfect people: They do not exist.”

District Cooling Systems

District Cooling Systems

District Cooling Systems – Many countries experience hot summers which impose a need for air conditioning in order to achieve and maintain a comfortable indoor environment. Air conditioning has traditionally been provided to buildings by electrically powered air conditioning units however, this equipment can have a high power demand. Many cities across the world are exploring alternative and more energy efficient cooling solutions.

One such solution is a District Cooling System (DCS) which involves the production and distribution of chilled water to multiple buildings through a network of insulated underground pipelines. Each building connected to this central system is connected by an Energy Transfer Station (ETS). The ETS typically uses a heat exchanger to chill down the water of the building using the district mains. This chilled water within the secondary building system is then used for space comfort cooling and process cooling. This essentially means that there is no requirement for chilled water plant to be installed locally in each building. Each building operator therefore purchases chilled water from the central provider.

District Cooling Systems include central chiller plant, pumps, valves & distribution pipework. Chilled water is often generated at the central plant station by dedicated chiller units. Free cooling can also be achieved by using water from nearby rivers, deep lakes etc. There are many examples of such systems such as in Paris where water from the river Seine is refrigerated to around 5oC and transported 71km to serve circa. 5 million square meters of offices, hotels, theatres, government buildings and of course, the Louvre. Another notable example case is one of the world’s largest district-cooling system schemes at The Pearl-Qatar.

Figure 1 provides an example of a District Cooling System as designed and modeled using FluidFlow software. This system uses two “central” cooling plant locations to provide chilled water which is then distributed through a network of 17.8 km pipework to serve 48 Energy Transfer Station’s.

District Cooling Systems

Figure 1: 505MW District Cooling System.

District Cooling Systems have been quoted as offering a more energy-efficient approach to achieving cooling as they can reportedly consume 20-35% less electricity in comparison to Individual water-cooled air conditioning systems and traditional air-cooled air conditioning systems respectively.

Other benefits of such systems include; lower CO2 emissions, reduced pollution, air-conditioning provided from a sustainable source, reduced maintenance costs for users, lower operating costs for users, increased reliability and increasing the potential for economics of scale. 

Fire Sprinkler Systems Technical Paper

Fire sprinkler systems can be found in a wide range of buildings or structures from warehouses to multi-story buildings, ships, oil rigs etc. Sprinkler systems are used to provide a reliable method of automatically detecting and extinguishing fires at an early stage. Many countries or regions have their own specific set of standards for sprinkler system design and installation. It is therefore always necessary to establish at the project outset, what specific set of regulations or standards must be adhered to when designing a sprinkler system. This is necessary in order to meet the approval of the local authority and indeed any specific insurance requirements. Typical standards used over the years include those issued by the Loss Prevention Council (LPC), British Standards Institution, Factory Mutual Insurance and of course & NFPA (National Fire Protection Association). NFPA Standard 13 serves as a benchmark for the design and installation of fire sprinkler systems.

The function of a building or type of work carried out within a building determines the fire risk and the design criteria for a sprinkler system. It is therefore imperative to discuss the design criteria for any given building with the local authority and buildings insurers prior to commencing design.

When designing sprinkler systems for fixed building structures, the local water authority should also be consulted at the preliminary design stage to obtain the relevant information regarding the suitability of the local water supply for supplying the system. In most cases, the local water main cannot be used to feed the sprinkler installation directly and as such, a fixed storage tank and pumping station may be required on site. The pumping station typically is made up of a duty and standby pump arrangement. Either pump must be capable of providing the performance characteristics required to suit the design system hydraulic calculations.

In the event of a fire in a building, once the sprinkler system has detected the fire, the water must be discharged at the required flow rate and spray pattern over the area of protection. The amount of water discharged from a sprinkler can be expressed using the following formula:

Q = K√P


Q is the water discharged in litre/minute.

K is a constant for the sprinkler.

P is the pressure at the sprinkler in bar.

The shape of the sprinkler deflector determines the spray pattern, water droplet size and area of coverage. There are a wide range of sprinklers designed for different purposes. It is therefore essential that the correct sprinkler is selected for the application under consideration.

As mentioned earlier, the function of a building determines the type of sprinkler system required including the spacing of the sprinkler heads required to ensure adequate coverage of the system in the event of fire conditions.

The spacing of sprinkler heads is largely dependent on the hazard type or hazard classification within the building. NFPA Standard 13 sets out spacing requirements which requires careful consideration when designing any sprinkler system.

The NFPA standard provides an example case of a large warehouse facility measuring approximately 200 ft x 130 ft. The building has approximately 200 sprinkler heads and the example case focused on 12 remote sprinklers in operation (see sprinklers within the blue dashed line in Figure 1). The system operates with a design flow rate of 259.6 GPM.

An outline plan and elevation of the building is shown in Figure 1.


Figure 1: Building Plan & Elevation.

Based on this operating condition, the system pipe friction loss calculations have been performed using the Hazen-Williams formula, as follows;image


P is the frictional resistance of the pipe (psi/ft).

Q is the flow in GPM.

C is the friction loss coefficient.

d is the internal diameter of the pipe in inches.

The Hazen-Williams Coefficient used in this worked example is C = 120 with the exception of the underground pipework which has a C = 150. The solution provided within the documentation utilises the equivalent pipe length method to estimate the loss associated with fittings in the system.

A key part of the completion of design calculations for any sprinkler system is the definition of the sprinkler coefficient, K value. This example utilizes a K factor of 5.6 for each sprinkler.

As outlined earlier, the example case focuses on 12 remote sprinklers in operation. The system was modeled in FluidFlow generating results which are a close match to those in the NFPA hand calculation. We can, however, expect differences between the two cases which can be attributed to the simplifying assumptions of using the equivalent pipe length method to determine the pressure loss across fittings. Note, FluidFlow doesn’t make this assumption and calculates the pressure loss according to the fitting characteristics. If for a specific design you have a requirement to use the equivalent pipe length approach, this method is also available in FluidFlow.


Figure 2: NFPA Sprinkler System Model.

The hand calculation notes a flow rate of 19.5 GPM serving the most remote nozzle whereas FluidFlow has determined a flow rate of 19.7 GPM. The static pressure calculated by the software is 12.1 psig which is an exact match of the hand calculation. The results for the branch pipes are noted as 40.2, 62.1 GPM in the hand calculation which again, compare well to the software result of GPM 40.2 & 61.3 respectively.

The minimum operating pressure for each sprinkler has been noted as 7 psi (0.5 bar). This minimum setpoint can be defined in FluidFlow and as such, any sprinklers which operate below this minimum pressure will be highlighted and clearly identifiable.

Sprinkler systems can be solved for any number or sequence of sprinklers activated at any given time. This solution can also be automated using the Scripting Module (Dynamic Analysis).

Using the pre-installed script option, it was established that this entire system has circa. 2421 ft of pipework with a total pipe volume of around 1.03m3.

The next example case is for a yarn processing facility (Figure 3). The design system flow rate, in this case, is 1639.9 l/min. The design is based on STD steel piping with a Hazen Williams pipe coefficient of 120. Each sprinkler head has an associated K factor of 115.



Figure 3: Yarn Facility Sprinkler System Model.

This system was solved with just 15 remote sprinkler nozzles in operation. As can be seen in Figure 3, all other nozzles are turned off. The flow rate in the most remote nozzle is calculated as 81.5 l/min and the associated pressure at the system inlet based on the defined design flow rate is 5.5 barg.


  1. NFPA 13 Standard for the Installation of Sprinkler Systems.
  2. Building Services Pocket Book (Second Edition) J Knight & P Jones.

Control Valve Sizing

Control Valve Sizing. The successful operation of process plant and systems involves the highest level of measurement and control performance. Control valves therefore perform a key function in this process. The performance of control valves can have a dramatic effect on plant operating efficiency, overall profitability and asset life cycle costs.

A correctly sized control valve can provide significant quantifiable savings as well as increase process availability, reduce process variability and reduce maintenance costs. Correctly sized control valves also last longer in comparison to unmatched or incorrectly sized valves. Careful consideration should therefore be given to the correct sizing of system control valves.

The Importance of Sizing

Control valve sizing procedures are based on accepted mathematical methods such as those detailed in ISA-75.01.01-2007 – Flow Equations for Sizing Control Valves. These methods can be used to develop accurate valve sizes however, during the course of plant design, control valves are often sized based on a future maximum design process load plus a safety factor. This can result in the specification, procurement and maintenance of a larger valve than necessary thus producing imprecise control, poor production performance in addition to the operational issues discussed above.

When sizing a control valve, the most common approach is to calculate the flow coefficient, Cv which is a measure of the capacity of the valve body and trim. A valve Cv can be described as the number of gallons per minute (GPM) at 60oF (15.5oC) that will pass through a valve with an associated pressure drop of 1 psi. In simple terms, a fully open control valve with a Cv of 14 usgpm/psi passes 14 usgpm of fluid with 1 psi pressure drop. The flow coefficient is discussed in further detail below.

Flow Coefficient (Cv)

The flow coefficient or valve coefficient denoted by “Cv”, is used to determine the valve size that will best allow the valve to pass the required flow rate while providing stable control of the process fluid. Most valve manufacturers publish Cv data in product catalogues for various valve styles. Table 1 provides an example of an equal percentage valve curve characteristic as published by Crane Engineering.

Table 1 Crane curve data points

 Table 1: Equal Percentage Characteristic Curve Co-ordinates

The performance curve for the tabulated co-ordinates is shown in Figure 1.

Figure 1 crane curve

Figure 1: Equal Percentage Characteristic Curve

Crane Engineering noted a value of 50% capacity at 82.3% travel which is denoted by the dashed line in the capacity curve. 

The tabulated coordinates for this performance curve have been defined in the FluidFlow control valve database. The control valve performance curve is automatically generated by the software and can be viewed in Figure 2.

Figure 2 crane equal curve DB

Figure 2: Control Valve Performance Curve (FluidFlow)

In an attempt to make a comparison with the Crane Engineering data, a test-case piping system was created in FluidFlow piping system software (Figure 3).

Figure 3 crane eq model

Figure 3: Crane Equal Percentage Control Valve Modeled in FluidFlow Software.

The control valve has been modeled in the above liquid flow system whereby the duty point is near the Crane Engineering case, 50% capacity at 82.3% travel. As shown in Figure 3, the calculated FluidFlow result matches exactly with the Crane Engineering example.

If the Cv for a control valve is not calculated correctly or accurately, the resultant selected valve will experience diminished performance. If the Cv is too small for the process, the valve itself or the trim inside the valve will be undersized resulting in the system being “starved” of the process fluid. Undersized valves exhibit a higher pressure drop across the valve to maintain adequate flow and exhibit limited flow capacity. Furthermore, since the restriction in the valve can cause a build-up of upstream pressure, higher back pressures created before the valve can lead to damage in upstream pumps or other upstream equipment. Although relatively uncommon, an undersized control valve generally cannot deliver sufficient flow rates under maximum load conditions.

If the Cv calculated is too high for the system requirements, the result is a larger oversized control valve is usually selected. In addition to the more obvious issues with an oversized valve such as a larger cost, weight and size, when in throttling service, significant control instability can occur. Usually the closure element, such as the valve plug or disk, is positioned just off the valve seat which leads to higher pressure drops across the valve and higher fluid velocities which can cause cavitation, flashing or erosion of the valve trim elements.

Cavitation can occur in liquid systems when high velocity reduces the static pressure inside the valve to below the pressure level at which the liquid stats to boil and produce vapor bubbles. These vapor bubbles collapse whenever the downstream pressure is higher than the vapor pressure causing high pressure waves. These implosions result in very high noise levels and can cause considerable damage to the valve body or trim parts under prolonged service.

Oversized control valves are quite common and are highly sensitive to operating conditions with even the smallest of adjustments in valve position causing significant changes in fluid flow rate. It is therefore extremely difficult to achieve the exact flow rate required under these conditions. Oversizing of control valves can also have a domino effect. Safety relief valves must be sized to match the capacity of the control valve. Within bypass configurations, isolation valves, bypass valves and drain valves must all be larger which can impact on the size of the piping and associated structural supports.

The most common and basic form of the equation for liquid applications is;



Cv is the Flow Coefficient which describes how much fluid will flow (GPM) through the valve for a given pressure drop of 1 psi.

Q is the Flow Rate (GPM).

P1 is the Upstream Pressure (psi).

P2 is the Downstream Pressure (psi).

SG is the liquid specific gravity.

The ISA-75.01.01-2007 standard includes equations for predicting the flow coefficient of compressible and incompressible fluids through control valves. The standard uses the following flow sizing equation for Newtonian liquid flow;

C = Q/N1 * √ (ρ10) / (∆P)


C is the Flow Coefficient which describes how much fluid will flow through the valve for a given pressure drop.

Q is the Flow Rate (m3/h, GPM, SCFH).

∆P is the differential pressure between upstream and downstream pressure taps (kPa, bar, psi).

ρ10 is the relative density.

N1 is the numerical units constant (Refer to Table 1 in ISA-75.01.01-2007 standard).

These equations are widely accepted for sizing valves in liquid flow systems.

Note, the equations detailed in the ISA-75.01.01-2007 standard enable the control valve size to be determined in absence of any connected piping and fittings. An even better solution is to apply these equations whilst also including the interactions of the pipework and fittings in a full connected piping system. This will provide a much more accurate solution and provide a baseline for correct valve selection.

Control Valve Sizing – Rules of Thumb

When sizing control valves, a general rule of thumb noted in many engineering publications is to size the valve such that it operates between 20 to 80% open at maximum required flow rate. It is also recommended to have the minimum opening no less than 20% to provide a safety margin at the minimum flow rate required. This approach ensures that as much of the valves control range as possible is used while maintaining a reasonable (but not excessive) safety margin.

As a guide, correctly sized globe valves are typically one size smaller than the line size and properly sized butterfly, full-ball and segment-ball valves are typically two sizes smaller than the line size. Note, this statement should be used as an indicative guide only and not as design criteria or a design rule. Furthermore, this guide only serves as a useful rule of thumb if the connected lines have been sized accurately and correctly.

Control Valve Sizing Example

The ISA-75.01.01-2007 standard provides an example liquid case which is summarised as follows;

Fluid: Water.

Inlet Temperature: 363K.

Fluid Density: 965.4 kg/m3.

Inlet Absolute Pressure: 680 kPa.

Outlet Absolute Pressure: 220 kPa.

Flow Rate (Q): 360 m3/h.

Pipe Size: 150mm.


C = Q/N1 * √ (ρ10) / (∆P)

C = 165 m3/h/bar for Kv

Modeling this scenario yields the following results.

Figure 4 ISA Liquid Example1

Figure 4: ISA Liquid Control Valve Sizing Example.

The control valve in this example has been automatically sized using FluidFlow software based on the design conditions presented. The modeled case provides a C or Kv of 165 m3/h/bar which matches the ISA-75.01.01-2007 example.

Energy Saving Opportunities

Control valve inefficiencies in plant processes offer opportunities for energy savings and reduced maintenance costs. Valves which consume a large fraction of the total pressure drop for the system or are excessively throttled can present considerable opportunities for energy savings. Pressure drops in liquid systems increase the energy requirements of these systems. Pressure drops are caused by resistance or friction in piping and in pipe bends, joints etc as well as throttled control valves. The power required to overcome the pressure drop is proportional to both the fluid flow rate and magnitude of the pressure drop. If the valve is oversized, the valve will be throttled excessively and also, if the valve is undersized, the pressure drop will be unnecessarily excessive which of course, can increase pump energy requirements significantly. Control valves should therefore be carefully sized and selected.

There is also scope for energy savings in existing systems featuring control valves. When reviewing design control valve sizes, it is important to understand the causes of any valve sizing errors. Emerson research previously identified several major contributing factors to valve sizing errors. These include selecting line-size valves, out of date process data resulting from changes in process conditions or operating conditions which differ to the original system design. To correct systems with incorrect valve sizes, its essential to obtain accurate process data at all expected operating conditions. The control valve can then be sized correctly.


The process of “control valve sizing” is a procedure where the dynamics of the system are matched to the performance characteristics of the valve. This produces a control valve of an appropriate size and type that will best meet the needs of managing flow within the process system. Every attempt should be made to carefully and accurately size and select a control valve for the required application. The topic of control valve selection will be discussed in a separate blog.

This discussion attempts to outline only the basics of control valve sizing. Further detailed reading is recommended when completing control valve sizing for liquid or even gas flow systems along with consideration of other design factors which should be considered as part of the sizing process such as Liquid Pressure Recovery Factor (FL), etc.

Note, FluidFlow solves the various forms of the equations described in ISA-75.01.01-2007, taking into account Liquid Pressure Recovery Factor (FL) etc.



An Insider’s Guide to Valve Sizing & Selection by Jon F. Monsen.

ISA-75.01.01-2007 Flow Equations for Sizing Control Valves.

Valve Handbook, Philip L. Skousen.

The Definitive Guide to Control Valves, Crane Engineering.

Closed Loop Systems

In general, there are two main types of systems in which pumps can be installed, open-loop and closed loop systems. Open-loop systems are circuits in which the pumped fluid is exposed to the local atmosphere at some point in the circuit. A typical open-loop system would be a cooling tower system where the flowpath is linear, i.e. fluid transfer between two vessels. Conversely, closed loop systems as the title suggests are closed piping circuits where the pumped fluid circulates in a closed loop without any exposure to the local environment and typically without transfer of fluid into or out of the closed loop. Examples of closed loop systems include hot oil circuits, cooling/chilled water systems, hot water heating and air conditioning systems. Figure 1 provides an illustration of a closed loop fresh water cooling system consisting of heat exchangers, circulating pumps, orifice plates and over 300 M of pipework.

Figure 1

Figure 1: Closed Loop Fresh Water Cooling

One of the unique aspects of closed loop piping systems is that the static elevation is not accounted for in head-pressure calculations as these systems are largely unaffected by static pressure. However, just like open flow systems, we still need to make checks for sufficient NPSHa, that the static pressure throughout the system does not fall below the fluid vapor pressure and therefore induce cavitation etc. Any pump selected for the closed loop system, must be able to transport the fluid to the highest point without flashing or inducing vacuum and the lowest point must also be evaluated for pump shut-off pressure. A closed loop circuit will exhibit only friction losses. Pumps operating in closed loop systems are therefore only required to overcome dynamic friction losses.

Let’s consider the operating conditions of a closed loop system with an elevation change of say, 10.0 M. The system circulating pump is required to transport the fluid from the bottom of the system (0 M) to the top at 10.0 M. It would initially appear that the pump must overcome the height difference of 10.0M however, due to gravitational effects, this isn’t the case as for every metre of fluid pumped vertically upwards, a corresponding 1.0 M of fluid drops on the return side of the system.

When the system is motionless, i.e. the fluid is not being circulated, the pumps suction and discharge exhibit the same pressure exerted by the two separate 10.0 M columns of fluid which are connected at the top. An alternative way of considering this is, the suction pressure available at the pump is equal to the discharge pressure required to move the fluid to the top of the system.

Regardless of where the pump is positioned in the loop, the differential head developed by the pump will always be the same.

As you would expect, the head required to maintain flow in a closed loop system decreases as flow decreases and becomes greater as flow increases. In well-designed systems, the friction losses will decrease proportionally with a decrease in flow rate.

Large scale closed loop systems can become complex to manually design as they often have many multiple branch lines or sub-loops.

When selecting a constant speed centrifugal pump for closed loop systems, the best efficiency point (BEP) on the pump efficiency curve should fall between the design minimum and maximum flow points on the pump capacity curve. This ensures the pump operating efficiency is maximised across the expected pump operating conditions.

The performance curves shown in Figure 2 show that, the centrifugal pump achieves the maximum flow condition of 1400 m3/h at 55.0 M TDH and rises just 5.0 M at the minimum design flow rate of 700 m3/h. Note, the flow experienced in this sample system at any point in time is based on the demand of the system.

Figure 2

Figure 2: Closed Loop – Constant Speed Pump

Note, a “flat” pump capacity curve profile is preferred for closed loop systems with variable flow due to the energy savings which can be achieved at lower flow conditions.

Although choosing a centrifugal pump with a “flat” capacity curve profile offers potential energy savings, much higher energy savings can be achieved by selecting a suitable VFD controlled pump (Figure 3). Since the system resistance curve reduces steadily from maximum to minimum flow (in this case 1400 to 700 m3/h), frequency control can be used to achieve the desired operating condition at low flow and in doing so, achieve much higher savings in comparison to that of a constant speed pump using valve throttling to adjust or regulate flow.

Figure 3: Closed-Loop – VFD Controlled Pump

The reason a much higher reduction in power is achieved by using the VFD pump is, a significant reduction in pump operating speed is achieved and secondly, the pump BEP efficiency isomer closely follows the system curve. As such, at 700 m3/h, the control frequency at the lower operating speed has almost the same efficiency as it does when the pump is operating at the speed required to achieve maximum flow.

If you are considering a VFD controlled centrifugal pump for an application, select the highest efficiency pump with a BEP that falls at or just to the left of the maximum flow condition. Keep head rise to minimum flow as low as the application will allow and test your pump selection using an appropriate software tool. This will allow you to evaluate the pump performance at different operating conditions.


The Practical Pumping Handbook, Ross Mackay.

Pump Selection for VFD Operation, Joe Evans.


Maximizing Steam Plant Performance Technical Paper

What are the different types of steam?

When water is heated beyond its boiling point it vaporizes into the gaseous phase or as we more commonly know it, steam. The properties of steam vary considerably and are greatly dependent on temperature and pressure.

Before we consider why we use steam, lets firstly consider the different types of steam.

Saturated Steam (Dry Steam): This form of steam occurs when water undergoes sensible heating to raise the fluid to its boiling point and then vaporizes with additional latent heating. Heating this steam further, i.e. above the saturation point, produces superheated steam.

Superheated Steam: Steam in this form occurs when saturated or unsaturated steam is heated beyond the saturated steam point. The result is steam at a higher temperature and lower density by comparison to saturated steam at the same pressure.

Unsaturated Steam (Wet Steam): This is probably the most common form of steam which is usually generated by steam boiler plant. It often contains non-vaporized water molecules which are carried over to the steam being transported throughout the piping distribution network. These water molecules can affect the performance of the steam plant and as such, it is for this reason steam systems are fitted with condensate removal equipment.

Why do we use Steam?

In its simplest terms, steam offers an excellent means of transferring a large mass of heat energy, usually via a piping distribution network. In terms of the operation of steam distribution systems, it’s the heat energy, pressure, temperature and flow rate of steam at each demand point which is of importance. This therefore places key emphasis on the design and operating performance of the piping distribution system.

Steam for power plant applications is often required at relatively high pressures and frequently needs to be superheated (dry gas). The presence of superheat improves the thermodynamic efficiency of the turbine operating cycle and also minimises condensation which can lead to erosion, leaking joints and glands. This in turn reduces the overall mechanical efficiency of the power plant. In general terms, a good system design will be based on a steam operating condition which yields the lowest pressure that will provide the heat output required, saturated and as dry as practicably possible.

An effective steam plant system design will produce steam at the desired quality (as dry as possible at the point of demand), will ensure the steam isn’t required to do more work than necessary, will be based on the process plant and equipment having a large enough heating area and ensure that the effective heat transfer from the steam to the process takes place at the highest practicable rate. The overall design can be enhanced further by ensuring that any heat given off from the process itself is harvested and utilised elsewhere on perhaps another application.

Can we Reduce Steam Demand?

Surprisingly, a considerable amount of heat energy is wasted in steam being required to carry out more work than necessary. When reviewing your steam plant, perhaps a useful starting point is to consider all components in the system between the steam generator and the demand point(s) which unnecessarily add to the total system heat demand. In general, when designing new steam systems, a good design approach is to consolidate all equipment with a steam requirement to one area and locate this area as close as possible to the steam main and ideally, steam generation plant.

There are a number of areas which can give rise to a reduction in steam demand. For instance, any steam pipework becoming redundant should be isolated or disconnected from the steam main as having redundant lines is particularly wasteful of heat energy, particularly if the piping is unlagged and/or exhibits signs of leakage. Let’s consider steam pipe sizes. If a steam pipe is oversized given the steam flow rate required, it becomes a continuous point of energy wastage. If the insulation on a steam pipe is in poor condition, has been applied incorrectly or is of a poor quality, the pipe will incur increased radiation losses which essentially means that some of the heat energy inherent in the steam line is simply wasted. A greater volume of condensate will be formed due to the greater heat loss which means wet steam will be delivered to the point of demand. To overcome this, additional steam trapping is required. Oversized lines are also by their nature, more expensive to install owing to the cost of the larger pipe, relatively higher cost of piping supports, fittings valves, insulation, and labour. These larger pipes can also contribute to a lower quality of steam due to the formation of additional condensate in the lines as noted earlier.

If steam piping is undersized, the system will incur higher pressure drops and insufficient steam flow to the demand point(s). There is also a higher risk of water hammer, noise and erosion. The enemy of the steam plant, water hammer, can cause damage to steam pipes, equipment and personnel and thus the likelihood of this phenomenon should be given due consideration and be eradicated from the system.

In older steam system installations, there can be bypass arrangements installed at steam traps. These bypass connections are often unnecessary and left open giving rise to significant heat loss. Furthermore, the correct and strategic installation of steam traps with appropriate air venting can eliminate the requirement for these bypasses and thus, the associated energy wastage.

In general, the lowest permissible operating temperature should be selected for a steam system. This obviously will optimize the amount of fuel required to generate the steam and maintain the design operating temperatures.

Giving due consideration to the points discussed, the overall system heat demand can be optimized to ensure the system is operating as effectively and efficiently as possible and consequently, reduce the overall site steam demand and associated plant fuel costs.

Can the Steam Distribution be Improved?

The dryness and quality of steam delivered to any demand point in a steam system depend heavily on how the steam is generated and of course, distributed throughout the site. In steam systems which are required to generate saturated steam, the steam begins to condense the moment it begins its journey through the distribution pipework. In general, the wetter the steam, the lower the quality for downstream process heating requirements. In such scenarios, steam carrying moisture can create a water film on the inside pipe surface. At this point it’s worth noting that, the water film is a very poor conductor of heat which of course, can inhibit the performance of the installation.

The steam distribution system can be improved by ensuring there is a correctly sized steam trap immediately downstream of the boiler plant. This trap will collect condensate from the lines thus improving the overall quality of the steam and improve efficiency.

Another point for consideration is the level of insulation fitted to the distribution pipework. Poorly insulated steam lines give rise to the formation of condensate in the steam, hence reducing the steam quality. The steam lines and in particular, flange connections and valves should be properly insulated to maximise the steam quality.

When selecting the insulation thickness for distribution pipework, the optimum insulation thickness is the economic thickness. Economic thickness considers the installed cost (material & labour) of the insulation as well as the ongoing value of energy savings over the lifetime of the plant. It is essentially defined as the thickness of insulation that minimizes the total lifecycle cost.

The following graph illustrates the lost energy cost over the lifespan of the system which decreases as the insulation thickness increases. The total cost curve relationship represents the sum of the installed insulation cost and the cost of the lost energy. The total cost curve shows a minimum value at the apex which represents the economic thickness of insulation.


Economic Thickness of Insulation (FluidFlow)

Overall, any water/condensate which develops in the main infrastructure pipework should be collected as early as possible at system low points.

In existing steam installations, it is very often the case that the installed pipework becomes too small or too large owing to changing usage patterns or changing demands imposed on the system which can develop and change over the years since the plant was originally installed. As a rule, pipes which are too small can result in lower than required steam flow at the demand point and also, pipes which are too large cause excessive heat loss due to the larger than necessary surface area/radiation surface.

A particular issue for consideration is elements of the piping system which have sagged due to lack of adequate piping supports. This sagging can result in water lodging at the base of the pipeline at the lowest point. This water can be a source of water hammer in a steam system which can be detrimental to the operation of the plant. Also, the passing steam can collect moisture from the water-logged point which if not removed by steam traps downstream, will be delivered to the demand point and once again, lower the quality of the delivered steam.

Another area of concern is the presence of air in the steam lines. Air gets into the lines when the boiler plant is off. This air should be vented from the distribution mains to improve the transfer of steam through the system. The air can be removed through the use of appropriately positioned thermostatic air vents. As a minimum, these vents should be fitted at terminal ends of various pipe runs.

It was noted earlier that the water film on the surface of the pipe is a poor conductor of heat. The air film however much more drastic. In fact, it is widely documented that air is over 1,500 times more resistant to heat transfer than steel piping material. It stands to reason that both the water and air films should be eliminated from the piping distribution system as rapidly as possible.

Steam Condensate

Another point of notable interest is the condensate and its recovery from the steam system. Condensate is essentially purified/distilled water which usually includes chemical treatment – ideal for use a boiler feed water. Condensate is of high monetary value owing to its inherent heat content and the fact it is of high quality, i.e. purified hot water. It should therefore be harvested at every available opportunity in the steam process. After all, it is much more cost effective to re-heat hot condensate into steam than it is to heat cold make up water into steam. In fact, condensate can be almost one third of the cost of generating steam. Take Note !

When developing or designing a condensate system, it should be borne in mind that condensate is a two-phase fluid and as such, sizing of the condensate lines is a much more complex process. In such cases, the fluid velocity requires careful consideration. Traditionally, condensate lines were sized on the basis they were liquid carrying pipelines. However, as we now know, these lines can carry both liquid and flash steam.

So what can I do with the recovered condensate? Firstly, we have seen that it is perfect as boiler feed water. However, it can also be used for space heating systems. The scenario which suits your site may be dictated based on your specific set of site conditions and requirements.

Giving consideration to the above, the performance and efficiency of a steam system can be radically improved through careful and considered recovery and usage of hot condensate.

CR System

Two-Phase Condensate Recovery System (FluidFlow)


This document attempts to outline some key points to be considered for when reviewing the performance and efficiencies of new and existing steam systems. Ensuring steam plant is consolidated to a specific zone, optimizing the amount of steam required including its temperature and pressure as well as its quality will increase plant efficiencies. Careful selection of pipe diameters will also have a considerable benefit in ensuring the plant operates effectively and efficiently.

A useful approach to take when reviewing the design and operation of any new or existing steam system is to think of the heat energy desperately trying to escape at every available opportunity! What can you do to prevent this phenomenon and develop a more effective and efficient system? Developing such a system will go a long way to reducing the overall fresh water demand, heat demand, associated fuel costs and site emissions.


  1. Steam & Energy Conservation – Spirax Sarco.
  2. The Proper Use of Steam – Spirax Sarco.

Hazen Williams vs Moody Friction Factor Pipeline Pressure Loss

This example case considers the flow of water in a 6 inch Schedule 40 pipeline and compares the results when using the Moody Friction Factor vs the Hazen Williams formula for the same flowing conditions. A hand calculation will initially be carried out to determine the pressure loss for the two scenarios. A model will then be developed to make a comparison with the hand calculation.

Problem Statement:

720 GPM of water at 95 F flows in a 6 inch Sch 40 steel pipe at a velocity of 8 ft/s. The pipeline has a total length of 1000 ft (equivalent length). Determine the friction loss using the Moody Friction Factor and the Hazen Williams formula.


Re = ρvD/μ


ρ               is taken as 62.057 lbm/ft3

μ               is taken as 0.000484 lbm/ft sec.

v                is 8 ft/s.

D               is 6.065 in.

Re = (62.057)(8)(6.065)/(0.000484)

Re = 5.2 x 105

The relative roughness of new steel pipe can be calculated as;

ϵ/D = (0.0002 ft / 6.065 in)(12 in/ft)

ϵ/D = 0.0004

Considering the Re of 5.2 x 105 and relative roughness of 0.0004, we arrive at a friction factor of 0.017 approximately.

The head loss due to friction can be determined as follows;

Hf = f (L/D) (V2 / 2g)

Hf = (0.017) (1000 ft / 6.065 in) (12 in/ft) (8ft/s)2 / (2(32.2 ft/s2)

Hf = 33.4 ft

Modeling this scenario yields the following results.


Figure 1: 6 Inch Pipe – Moody Friction Factor.

The modeled value for friction loss is 32.3 ft fluid which is close to the hand calculation value of 33.4 ft. The difference can be attributed to the simplifications of the hand calculation by comparison to the modeled solution.

Let’s now consider the Hazen Williams approach using the following equation:

H = Cf L Q1.852 / C1.852 D4.87


Cf              is the unit conversion fact (4.72).

L                is the pipe length (1000 ft).

Q               is the volumetric flow rate (8 ft/s)(0.20063 ft2) = 1.605 ft3/s.

C               is the Hazen Williams coefficient, initially considered to be 100 for steel pipe. Note, this is a conservative value and allows for future scaling of the internal pipe surface. The American Iron & Steel Institute’s Committee for Steel Pipe Producers recommends a C value of 140.

H = (4.72) (1000) (1.605)1.852 / (100)1.852 (0.505)4.87

H = 11337 / 181.6

H = 62.4 ft.

Modeling this scenario yields the following results.



Figure 2: 6 Inch Pipe – Hazen Williams (C = 100).

The modeled solution (62.1 ft fluid) compares well with the hand calculation results of 62.4 ft.

Note, if we use a C value of 140 for clean or new steel pipework, we get;

H = 11337 / (140)1.852 (0.505)4.87

H = 33.5 ft.

Modeling this scenario yields the following results.



Figure 3: 6 Inch Pipe – Hazen Williams (C = 140).

The modeled value for friction loss is 33.3 ft fluid which is close to the hand calculation value of 33.5 ft. Again, the difference can be attributed to the simplifications of the hand calculation by comparison to the modeled solution.

We have already calculated a friction loss of 33.4 ft using the Moody Friction Factor. Conversely, if we use a roughness value for corroded steel pipe (0.013 ft), we get;

ϵ/D = (0.013 ft / 6.065 in)(12 in/ft)

ϵ/D = 0.03

From the Moody diagram, the friction factor is approximately 0.0265. This increases the friction loss by a factor of 0.0265 / 0.017 which yields,

H = 33.4 ft (0.0265/0.017)

H = 52.1 ft

Review this scenario in a model yields the following results.



Figure 4: 6 Inch Pipe – Fixed Friction Factor of 0.0265.

Let’s summarise the results.

Screen Shot 2017-09-19 at 12.01.30

This study demonstrates the following;

  1. There is good agreement between the Darcy-Weisbach and Hazen Williams equations for clean pipework.
  2. The results are sensitive to the roughness and C factor values for corroded or old pipework installations.


  1. Piping Systems Manual (Brian Silowash).

Net Positive Suction Head (NPSH) Technical Paper

When considering Net Positive Suction Head, it is useful to differentiate between available net positive suction head (NPSHA) and required net positive suction head (NPSHR).

NPSHA is a characteristic of the system in which a centrifugal pump operates and represents the difference between the absolute suction head and the fluid vapor pressure at the prevailing temperature.

NPSHR is a function of the pump design and represents the minimum required margin between the suction head and fluid vapor pressure.

The way NPSHA is calculated depends on the system configuration. The following Figures help illustrate this point.

Figure 1 outlines how NPSHA is calculated for a given capacity of water at 80 F based on a system with suction lift (15ft).

Figure 1: NPSHA Calculation with Suction Lift.

The hand calculation of NPSHA for this scenario (at sea level) would be:

NPSHA = 2.31 (Ps – Pv) / SG + Z – Hf


Ps         is the pressure above the liquid surface (psia).

Pv         is the vapor pressure of the liquid (psia).

Z          is the static head (ft)

Hf         is the friction losses (ft). The friction losses total 2.38 + 0.46 + 0.16 = 3 ft.


NPSHA = 2.31 (Ps – Pv) / SG + Z – Hf

NPSHA = 2.31 (14.7 – 0.5) / 1.0 – 15 – 3

NPSHA = 14.8 ft.

The hand calculation produces a NPSHA of 14.8 ft which match the modeled result.

Let’s consider a scenario whereby the system inlet features a pump taking its suction from a pressurized tank at an elevation of 10ft. The fluid, in this case, is once again water at 80 F.


Figure 2: NPSHA Calculation with Suction from a pressurized tank.


The hand calculation of NPSHA for this scenario would be:

NPSHA = 2.31 (Ps – Pv) / SG + Z – Hf

NPSHA = 2.31 (14.7 + 5 – 0.5) / 1.0 + 10 – 4.001

NPSHA = 2.31 (14.7 + 5 – 0.5) / 1.0 + 10 – 4.001

NPSHA = 50.4 ft

The hand calculation produces a NPSHA of 50.4 ft which matches the modeled result.

Note, both NPSHA and NPSHR vary with capacity. With a given static pressure or elevation difference at the suction side of a centrifugal pump, NPSHA is reduced at larger flow rates by the friction losses in the suction piping and fittings. On the other hand, NPSHR being a function of the velocities in the pump suction passages and at the inlet of the impeller, increases basically as the square of the capacity. The NPSHR for a given pump is provided by a pump vendor. Many factors affect the estimation of NPSHR such as eye diameter, number of impeller vanes, suction area of the impeller, shape of the vanes, shaft and impeller hub diameter, impeller specific speed, shape of suction passages, etc. it is therefore not recommended that a designer estimates a value of NPSHR for a pump but obtain this specific data from the pump vendor.

The NPSHR describes the amount of pressure required at the inlet of a pump to prevent air bubbles from forming inside the pump unit. If a scenario arises whereby the NPSHA is lower than the NPSHR by the pump, air bubbles are allowed to form. These bubbles can implode violently inside the pump causing significant damage. This effect is known as cavitation.

Pump cavitation becomes evident when there is one or more of the following conditions present in a system; noise, vibration, drop in pump head-capacity and efficiency curves and with time, damage to the impeller by pitting and erosion.

A useful point to note is that NPSHR curves provided by pump manufacturer’s are usually based on using cold water for the pump test conditions. Thus, it might be assumed that the NPSHR by a centrifugal pump for satisfactory operation is independent of liquid vapor pressure at the pumping temperature and of course, this is not true. The NPSHR for a given capacity can vary appreciably for different fluids over a range of temperatures. In its simplest form, even when pumping water the NPSHR decreases when the water temperature increases.

What options are available to me to increase the NPSHA:

  1. Reduce the resistance in the suction side of the system and hence the associated pressure losses.
  2. Raise the height of the supply tank in open systems.
  3. Raise the liquid level in the supply tank.
  4. Lower the pump installation height.
  5. Increase the surface pressure of the liquid in a closed system.
  6. Monitor and control the fluid temperature.
  7. Lower the pump operating speed.
  8. Use a larger impeller eye area.
  9. Use several smaller pumps in parallel.

Item 9 above may initially appear to be a costly solution however, in many cases, three half capacity pumps of which one may be a spare, are often no more expensive than one full capacity pump plus its spare. In many cases, just two half-capacity pumps can be installed without a spare since part-load can still be carried if one pump has failed and is temporarily out of service. In addition, if the demand varies widely, operating a single pump during light-load conditions will save energy.

This technical paper attempts to briefly describe NPSH, outline some of problems which can arise when the NPSHA is lower than the NPSHR and what measures can be considered to remedy such conditions. In the end, careful design together with liaising with the pump vendor regarding the NPSH requirements for a proposed pump model will go a long way to eliminating any potential operational problems.


  1. CIBSE Guide B1 2016.

Petrojarl 1 FPSO (Marine Systems, Floating Production, Storage and Offloading Unit)

Back in November 2014, Nevesbu and Iv-Oil & Gas joined forces and started the Pre-Contract Engineering for the contract to modify and upgrade the Petrojarl 1 FPSO. It was anticipated that the upgrade will take around a year to complete. On completion of the project, the FPSO will be used as an early production system (EPS) unit on the Atlanta field located around 185 kilometres offshore from the Brazil coast. This project presented many challenges and was also unique in that an EPC redeployment upgrade contract had not been carried out in Europe before.

Teekay Offshore awarded Damen Shiprepair Rotterdam (DSR) the complete EPC contract for the modification and upgrade of the Petrojarl 1 FPSO. DSR in turn contracted Nevesbu to take care of the Marine Systems, Structural (design and engineering) and Naval Architecture for all modifications. Iv-Oil & Gas was then requested by Nevesbu to execute the Process, Mechanical, Electrical, Instrumentation, Structural, Piping Design and Piping Engineering. Frames Process Systems (FPS) was also selected for its expertise in the field of separation. They provided a large amount of the skid mounted package units (skids) and worked closely together with Nevesbu and Iv-Oil & Gas for integration of the skids with the vessel.

On completion of the initial demolition works, in April 2015, the Petrojarl 1 was moved into DSR’s dry dock, where the refurbishment, modification, and upgrade of the ship was carried out.

Fast-forward to August 18, 2017, Damen Shiprepair Rotterdam (DSR) successfully delivered the FPSO Petrojarl 1 to Teekay Offshore following a complete redeployment project which took place over the past 2 ½ years.




Scope of work 
The self-propelled FPSO spent 14 months in Dock No. 8 (300 x 50m) undergoing refurbishment of its marine systems, underwater hull, seawater system, crane booms, heating coils in the cargo tanks and specialised steelworks in the upper and lower turret areas, which needed to be completely revised and adapted to suit the 1500 metre deep mooring location. Simultaneously, new designed high quality, prefabricated equipment skids containing heating, cooling, separation, compression, boilers, centrifuges as well as a new E-house with electrical equipment were placed on board. Interconnecting piping and cabling was subsequently installed to complete the topsides and connect it to the remaining facilities.

The process installation consisted of the following:

  • Crude oil, produced water and associated gas separation;
  • Crude oil dehydration and desalting;
  • Produced water treatment;
  • Associated gas treatment and fuel gas distribution;
  • Fresh water washing system to meet the maximum allowable salt content in the stabilised crude oil;
  • Upgrade of existing seawater (cooling) system;
  • Steam system to heat the produced crude oil and generate fresh water from seawater;
  • Upgrade of chemical injection systems;
  • Upgrade of existing metering systems to comply with Brasilian requirements;
  • Pigging facilities to enable the launching/receiving of pigs from pigging operations;
  • Modify the existing piping arrangement on the manifold deck;
  • Upgrade of existing subsea controls;
  • Well services/diesel flush system to allow for diesel to be used for flushing crude oil from production, service risers or flowlines on shutdown, for preservation and also for heating prior to startup;
  • Upgrade of well kill system to reduce drilling rig workover time and to reduce produced crude oil handling in the rig using brine.

According to DSR, the project involved more than 450,000 engineering hours and more than 50% of the process equipment was removed and replaced by new and additional equipment.

Teekay had operated the Petrojarl 1 for 28 years in the North Sea, but it is now destined for the Atlanta field offshore Brazil.

Located in block BS-4 in the Santos basin, Atlanta is a postsalt oil field, 185 km (115 mi) offshore Rio de Janeiro in a water depth of 1,500 m (4,921 ft). First oil is expected in 1Q 2018.

Queiroz Galvão Exploração e Produção S.A. is the operator of the block with a 30% ownership along with consortium members OGX Petróleo e Gás S.A. (40%) and Barra Energia do Brasil Petróleo e Gás Ltda. (30%).

We are proud that FluidFlow software was used to design and model the upgraded seawater cooling for topsides and diesel transfer system of the Petrojarl 1 FPSO vessel together with checks on other piping systems required for the successful operation of the FPSO.



FluidFlow v3.43

General Release summary: The thermodynamic capabilities have been extended for single phase fluids. FluidFlow can now accurately cover all phase regions and enthalpy paths through and from supercritical phase region. Calculations within supercritical region are now valid. This release also contains some minor bug fixes.

General Release info:

* Extended the thermodynamic capabilities so that the product can accurately cover all regions and enthalpy paths through supercritical phase region. Calculations within the supercritical region are also available.

* Supercritical phase state now fully detected and shown in results.

* Added marine fuel oils to fluids database.

Valve Authority – Technical Paper

Control Valve Authority

Control Valve Authority – Technical Paper

Valve Authority is a term used to describe the basis on which a control valve is selected. Quality in design represents the correct consideration of many engineering decisions and sizing control valves based on authority is an important factor. If the control valve is oversized, the performance, life span and reliability of the valve and other equipment items will be reduced. If we undersize then we may have the authority but cannot achieve the design capacity. Additionally, designs should consider energy optimization but this often leads to poor authority.

The Valve Authority (N) is generally defined as the ratio of the pressure drop across the fully open valve compared to the pressure drop across the entire circuit (including the valve) at design flow conditions. Valve Authority is expressed using the following equation:

N =ΔPvalve /ΔPtotal


N              is the Valve Authority.

ΔPvalve    is the pressure drop across the valve in the fully open position.

ΔPtotal     is the total pressure drop across the circuit.

In order to develop good control, it is recommended that a control valve is selected to achieve a valve authority of 0.5 or greater. An authority below 0.25 gives unstable control; 0.25 – 0.5 gives fair to good control whereas 0.5 -1.0 gives excellent control. The higher the authority, the greater the energy wastage.

Let’s consider a simple water circuit example where the available pump pressure is 13 kPa at maximum flow conditions, i.e. the valve is in the fully open position. This 13 kPa represents the total frictional resistance of the circuit including the control valve which has a pressure drop of 6 kPa in the fully open position.

Valve Authority


Considering the pressure drop across the control valve, the pressure drop across the remainder of the circuit will be 13 – 6 = 7 kPa. The valve authority can therefore be calculated as follows;

N = 6 / 6 + 7

N = 0.46

Let’s consider a design example to explore the phenomenon of valve authority a little further.

Problem Statement:

It is intended to model a section of pipework downstream of a heat exchanger transporting Methyl Diethanolamine (MDEA) going to a column. The system pressure shall be controlled by a pressure reducing control valve with a design valve authority of approximately 0.5. The engineer is also required to ensure that the pipelines in the system are to be sized such that the gas superficial velocity (vapor velocity) exiting the system is between 10 & 15 m/s.

Design Data:

Temperature of MDEA at system inlet: 106.5 °C.

Vessel Surface Pressure: 5 bar a.

Elevation to Vessel Base: 3 m.

Outlet Elevation: 31.27 m.

Pressure at outlet: 1.8 bar a.

Design Flow at system outlet: 367,000 kg/h.

Fluid Vapor Quality Approx. 0.06.

Solution Approach:

Let’s develop this outline scheme design such that the pressure drop across the control valve is minimised which is consistent with good control. Good control of course means that the design valve authority should be around 0.5 and in all circumstances, must be above 0.2.

We know the vessel surface pressure is 5 bar a and the design outlet pressure is 1.8 bar a. The total system pressure loss between the inlet and outlet of the circuit can be established as 3.2 bar. This means that, based on a design valve authority of 0.5, we can estimate the valve pressure loss at the design flow rate of 367,000 kg/h to be around 1.6 bar (3.2 bar x 0.5 = 1.6 bar). This data can be used as the basis of a preliminary design as outlined in Step 1.

Step 1

A preliminary design of the system has been developed based on the design flow rate provided. A Flow Coefficient (Kv component) can be used in the first instance to represent a pressure reducing valve as at this stage, we simply wish to define a fixed pressure drop of 1.6 bar across the component. Figure 2 provides an illustration of the system.


MDEA Circuit

Figure 2: MDEA Circuit – Control Valve represented by a Flow Coefficient (Kv).

When this initial design is calculated, the MDEA in the system is detected as being a two-phase fluid with a vapor quality of approximately 6.7%. For two-phase systems, a useful rule of thumb is to keep the gas superficial velocity around 15 m/s at this vapor quality.

At this point, it’s often useful to know what flow regime is encountered in the system. In this particular design case, the flow in the lines is in the annular mist regime which is preferable to slug flow which of course should be avoided. This undesirable slug flow regime can be avoided by keeping the gas superficial velocity relatively high.

Annular Mist Flow Pattern Map

Figure 3: Annular Mist Flow Pattern Map.

Various pipe diameters can be applied to this system in an attempt to achieve a flowing velocity which is as near as possible to the design velocity of 10 to 15 m/s. This preliminary design case has been solved using 300mm pipes which generates an initial velocity of 14 m/s. It is worth noting at this point that as the fluid flows along the pipelines, the frictional resistance of the various pipes and fittings results in an increase in the vapor quality and hence, gas volumetric flow rate and flowing gas velocity. This phenomenon would at this early stage indicate that it may be prudent to increase the pipe diameters in the system – perhaps downstream of the pressure reducing valve. Indeed, a check on the velocity at the system outlet indicates the velocity increases to 112 m/s. This velocity is clearly much too high and should be reduced as the design evolves.

Step 2

The outline design completed in Step 1 can now be developed further. The Kv component can be changed to a PRV (pressure reducing valve) and this valve can be automatically sized. When sizing the valve, the outlet pressure in the MDEA system can be set to 1.8 bar a (based on the design data provided). The design setpoint pressure of the PRV can then be adjusted until the design flow rate of 367,000 kg/h is achieved in the system.


MDEA Circuit with Pressure Reducing Valve

Figure 4: MDEA Circuit with Pressure Reducing Valve (PRV).

Step 3

At this point, it is worth considering different pipe diameters downstream of the PRV. The PRV setpoints pressure can also be adjusted until we obtain the required design flow for each pipe diameter considered. The calculated results are outlined in Table 1.


valve authority PRV figures

Table 1: Review of Valve Authority for different Line Sizes & PRV Set-points.

We can see that the preliminary line sizes of 450 – 500 mm produces a sufficiently high valve authority (N = 0.525 & 0.56).

The next stage is to select the correct line size whilst giving consideration to the limiting velocity criteria of 15 m/s for the vapor phase. The most suitable pipe diameter is therefore the 500 mm pipe. Smaller pipe will also work, however we need to avoid choked flow in the system. We therefore need to keep the gas velocity below a max of 50 m/s. If capital cost was the main design criteria, you could consider using a smaller pipe size.

Figure 5 provides an illustration of the system at this stage of development.


guide to Valve Authority

Figure 5: MDEA Circuit – Pipe Sizes Optimised.

So far, the calculations have produced the valve Cv along with fluid physical properties. This data can be provided to the valve supplier so that they can make the final valve selection for their range of equipment.

Step 4
Finally, in this step the automatically sized pressure reducer can be swapped for the actual vendor control valve (Figure 6). In this final case, the pressure loss across the valve is calculated to be 1.5 bar which produces a valve authority of 0.47 (1.5/3.2 = 0.47).

Valve Authority

Figure 6: MDEA Circuit – Vendor Control Valve.

This example helps us consider the application of valve authority in a real system design whilst giving due consideration to other aspects of the system such as fluid physical properties, valve Cv, design flow rates, pressures etc.

FluidFlow v3.42

Fixed a bug that occured when selecting “mm Water g” OR “m Water g” units.This was originally fixed in V3.39 but regressed in V3.41. New procedures in place to eliminate code regressions.

FluidFlow v3.41

General Release info:

Improved pipe heat loss calculation with the addition of new correlation for estimating outside film heat transfer coefficient. Bug fixes.


* Crane Tee Junction TP410 pre 2009 relationships. New recommendation is to use for existing/legacy calculations ONLY. Input editor text now reflects this.

The relationships are just too simplistic to realistically predict the pressure losses over all possible operating conditions and totally ignore any pressure recovery effects.

* All Heat Transfer Coefficients in pipe results were shown as W/m C where units shown are W/m2 C. Coefficient values now normalised to W/m2C by multiplying original value by the log mean radius.

This is more in line with most literature sources. Change was made not because answers or calculations are incorrect but because it is now easier compare values directly with common literature sources.

* Following a suggestion from some of our German customers, we incorporated a new improved method for estimating the outside coefficient in pipe heat loss/gain calculations.

The literature source for the improved method is the Springer publication “VDI Heat Atlas (VDI Wärmeatlas)”.

Previously FluidFlow used ASTM Standard C680. It is still possible to use the original ASTM Standard C680 relationship, selected from the Options -> Calculations Dialog -> Global Settings tab.

* Co2 gas density and specific heat definitions have been adjusted to provide more accurate results over a wider temperature and pressure range.


* Diffusers at choked conditions show the inlet results as outlet results.

* In networks containing centrifugal pumps and more than one fluids with viscosities of > 800 cP, sometimes Pump viscosity corrections were made when this should not be the case.

* Outside heat loss convection was overestimated in rare cases for gas flow along medium/long pipes.

* Improved convergence testing to prevent the solver converging too soon. For gas calculations this sometimes resulted in incorrect downstream temperatures.

* Reservoir No Flow – Bug in creation of accumulator for solver.

* Pump derating and speed changes were not working when both effects occurred together.

FluidFlow v3.40

General Release info:

Control Valves: Improved consistency of calculated valve position and valve coefficient over the complete operating range.

Polyethylene pipes, new pipe sizes added.

Petroleum fraction properties, NBP050 to NPB450 range added to fluids database.


  • Added back a directional definition for all flow control valves. This requires the user to specify the flow direction through the valve and is necessary to reduce interaction between flow control valves in large networks
  • Removed Text Import and Export menu options because format is now out of date and is not supported in the future.


  • Multiple flow controllers in a network could occasionally cause convergence BEFORE the network has actually converged. See changes above.
  • Liquid line with heat transfer in buried pipes, sometimes calculated out phase state as two-phase when phase state should be liquid.
  • Pipe scaling was not read in properly from old files (pre V3.30).

FluidFlow v3.39

General Release info:

Calculation procedures for viscosity correction method for centrifugal pumps has been updated to HI 2015 guidelines

Bug Fix:

Units m Water g and mm Water g went missing in V3.38. Now reinstated.

FluidFlow v3.38

Contains fixes for all reported bugs up to end of September 2016.

General Release info:
No new enhancements, bug fix maintenance release only

None in this release

* Fixed a rounding error (<2%) that occurred when converting to m Water and m Water gauge.
* Improved the chart visibility of slug region in two-phase flow pattern maps, previously part of the area was overwritten by elongated bubble regime.
* Fixed a bug in outlet velocity calculation when phase change occurs within a pipe.
* Fixed a bug that caused phase change within a pipe when Heat Loss Model = “Ignore Heat Loss/Gain”, without taking into account heat of vaporization.
* Fixed a bug that caused flow reversal at known flow nodes out of a network at very high specified flows i.e. many times greater than sonic flow.
* Fortis Only – Overcome the instability caused by the discontinuities in the Universal Gas Sizing Equation for control valves.
* Fixed a bug which shows gauge pressure results incorrectly if atmospheric pressure is changed.

Two-Phase Flow


Gas-liquid two-phase pipe flow is of significant importance in a wide range of engineering industries such as steam generators, chemical process plant, distillation processes and heat transfer systems. The design of these systems is often a complex phenomenon

Using suitable engineering simulation software can help the engineer design efficient and effective systems, understand plant performance and quickly evaluate alternative design scenarios.

In two-phase flow, the vapor mass fraction is often not constant and there is mass transfer between the fluid phases. FluidFlow takes this into account in your model solution. In fact, you can see the results for inlet and outlet vapor quality for all pipes and elements in your system. Flow pattern maps are generated automatically for all pipe in the system, helping you identify flow regimes and any undesirable operating conditions.

FluidFlow is used successfully by engineers to calculate pressure losses and flow distribution in two-phase pipe flow systems. The simulation software will automatically track fluid phase-state throughout the piping distribution system and the software is provided with a comprehensive database of two-phase fluids, boosters and associated piping equipment. Automatic control valve and equipment sizing is included helping you to accelerate the design process.

FluidFlow is easy to use and new users are provided with a Designer Handbook meaning you can tackle those design projects instantly.

For more information on Two-Phase Flow click here

Compressible Flow Systems

compressible flow

FluidFlow customers use the software to design and optimize a wide range of compressible pipe flow systems including; natural gas transmission pipelines, steam distribution systems and compressed air systems.

For compressible fluid flow in pipes, the pressure and temperature conditions continuously change as a gas or vapor flows along a pipeline. This means that the physical properties of density, viscosity, heat capacity, thermal conductivity, velocity etc, change with pipe length.

FluidFlow uses a number of compressible flow equations, and incorporates the Joule Thomson effect to obtain a rigorous solution which is accurate for both low and high velocity flow systems.

By using FluidFlow, engineers can accurately calculate compressible flow through an orifice plate, control valve, relief device, nozzles, valves and all common piping components. You can also automatically sizing pipes, pumps, ducts, fans, compressors, control valves, relief devices (ISO & API), orifice plates and nozzles.

The software can also calculate heat loss/gain from pipes and model buried pipe heat transfer. The software is provided with a library of insulation materials as standard and engineers can select the desired insulation thickness. Convection, conduction and radiation losses are calculated. This means you can use FluidFlow to optimize energy use by selecting the economic insulation thickness.

For more information on Compressible Flow click here

FluidFlow Pressure Drop Calculator

FluidFlow is a pipe flow calculator which is used to perform fluid flow analysis in piping systems featuring heat exchangers, orifice plates, control valves, pumps, venturi flow meters and equipment items.

The software is easy to use and supported by an experienced team of engineers which are always willing to lend a helping hand.


FluidFlow is a modular pipe flow calculator

Available calculation modules include

  • Liquid
  • Gas
  • Two-Phase
  • Non Newtonian and Settling Slurry
  • Scripting (Dynamic Analysis)


FluidFlow Liquid Module

The FluidFlow Liquid Module is a water flow calculator which allows you to define vendor equipment to a database such as pumps, control valves, pumps, venturi flow meters etc. When defining pumps, you can enter your vendor-specific pump curves to the database which will be stored for all your modeling projects. You can then model the performance of the pump in your system.


The software enables designers to complete fluid flow simulation studies in an instant and automatically size equipment, taking the pain out of your design projects.


Flow Measurement In Piping Systems

FluidFlow is a flow calculator which allows you to understand flow measurement in your piping systems. You can input data obtained from a site pressure test to a model and understand exactly how your piping system is performing.


FluidFlow solves the continuity of mass, energy and momentum equations. You can clearly view result for flow rate, pressure, pressure loss, velocity, density, viscosity, temperature, Reynolds number and friction factor.

FluidFlow Overview

FluidFlow is primarily a maintenance release addressing reported bugs and adding new features requested by our users. It is recommended that all users upgrade to this release. We have updated our control valve calculation code to reflect the very latest Instrumentation Society America ISA-75 Guidelines, this includes choking detection for both liquids and gases.


  • Large networks containing many tee junctions now converge quicker. Note, it is still important to place the branch orientation (red dot) correctly.
  • Improvements made to equipment sizing consistency.
  • All Control Valve Equations now updated from ISA 1985 to ISA 2007 guide. For choked valves a warning is now provided. Added calculation of Cv at liquid choked flow conditions.
  • Added network data caching to improve performance of network version.
  • Added the ability to use visual elements on script forms (labels, list boxes, combo boxes, tabs, grids, etc.).


  • Open pipe exit pressure reverted back to pre 3.31 where stagnation pressure is assumed to be atmospheric.
  • Removed mouse wheel support for the input editor to prevent clashes with flowsheet occurring. Flowsheet now works better using mouse wheel.

FluidFlow v3.37

Primarily a release for Fortis, including all requested logging and database enhancements, bug fixes regarding gas reducers, and series choking.

General Release info:


  • Added the ability to automatically adjust atmospheric pressure for altitude.
    Options -> Calculation-> Global Settings.


  • Fixed an overwrite of value 101325 Pa a, for atmospheric pressure that sometimes occurred when resetting defaults.
  • Improved network convergence for multiple flow controllers.
  • New warning advising when a pump is acting as a turbine.
  • PD Pump auto-sizing bug fixed.
  • Fixed a bug which caused a program crash when changing to French language.

FluidFlow v3.33

Version 3.33 is primarily a maintenance release addressing reported bugs and adding new features requested by our users. It is recommended that all users upgrade to this release.

We have added the ability to make buried pipe heat loss calculations. To support this calculation, different pipe coatings, soil, and backfill types have been added to the Insulation database. In addition, the thermal conductivity as a function of temperature relationship for all insulation materials and soils has been improved.

For gases flowing at or near the saturation point in pipes with heat loss on, we have added an option to include condensate traps. If this option is “on”, the flow will stay in the gas phase at the saturation point and the software will report the mass of condensate to be removed. By default, this option is “off”, and so some of the gas will condense and become two-phase as it flows down the pipe. This means the vapor quality decreases along the flowpath because the condensate is not removed.


  • Added the ability to make a buried pipe calculation. Added different pipe coating, soil, and backfill types to the insulation database.
  • Added phosphine gas to physical property database.
  • Improved insulation thermal conductivity as a function of temperature relationship and added more data for insulation materials.
  • Added the ability to assume steam traps are present for a condensing gas. This option (‘Options | Calculation…’ menu item; Gas page) prevents gasflow from developing into 2-phase flow when heat loss is included.
  • Bill of Materials now subdivides pipes into schedules.
  • Added ability to model expansion loops in one component, instead of drawing out each individual expansion loop.
  • Added more calculation examples to QA tests and updated help file.


  • For Open Pipes and Open Boundaries with a resistance, the exit static pressure is now assumed to be atmospheric pressure. In earlier releases, the exit stagnation pressure was assumed to be atmospheric.
  • Restricted tee junction K values to maximum and minimum values: Max = 90 and Min = -15. This aids convergence without limiting practical values.
  • Added the ability to include Joule Thomson Coefficient in “Do Heat Loss Calculation”, via the ‘Options | Calculation…’ menu item; Gas page.

FluidFlow v3.31

With version 3.31, we have really focused on empowering you to build your models faster. We’ve implemented unique, cutting edge tools that will not only reduce the time it takes you to build your model on the flowsheet – we’ll also do the design for you!

Automatic Pipe Sizing: In addition to the economic pipe sizing feature, designers can now specify a desired pressure gradient or nominal velocity for each pipe element in the model. FluidFlow will then calculate the pipe diameter required to achieve that design constraint

Automatic Equipment Sizing: We already have auto-sizing in place for safety relief valves and burst disks to both API & ISO standards for liquid, gas, steam and two-phase flow systems. Now FluidFlow can auto-size all your key equipment items. Pumps, compressors, fans, orifice plates, nozzles, venturi tubes, pressure and flow controllers can now all be sized automatically, saving you time and effort.

Automatic Flow Balancing with Orifice Plates: Orifice plates can now be sized based on design pressure loss or flow rates. No more iteration in your design as you adjust orifice diameter to achieve your desired flowrate – FluidFlow will do it all for you.

Flowsheet Improvements: Getting your flowsheet built quicker is key to efficient modelling. So we’ve changed our directional components. No more “red dot” to define the flow direction of a controller. Just drop the pump or controller on the flowsheet and FluidFlow will work it all out for you. You can now also hold down the CTRL key when adding a template to continue adding multiples of the same template.

Reporting Upgrades You can now view your flowsheet data report from directly within FluidFlow. No need to export to excel to look at the raw data. The report has its own dialog box, so you can put it on a separate screen, or have it side by side with your flowsheet. We’ve improved the print quality of our exports too, so your .pdf exports are now crystal clear.


  • Ability to autosize the following equipment items: Pipes, Centrifugal Pumps, Compressors, Fans, PD Pumps, Pressure Reducers, Pressure Sustainers, Differential Pressure Controllers, Flow Controllers, Orifice Plates, Inline Nozzles, Venturi Tubes, Safety Relief Valves, and Bursting Disks.
  • Added 2 additional pipe sizing criteria options: Pressure Gradient and a user-entered Velocity.
  • No longer necessary to define the flow direction of pressure controllers.
  • No longer necessary to define the flow direction of flow controllers.
  • Hold down CTRL when Inserting a Template to continue insertion. Same logic applied to Paste.
  • Components Bar (‘View | Components Bar’)
  • New Data Report. Flowsheet Data/Data Report. (‘View | Flowsheet Data’)
  • Improved Print and Export Quality.
  • Integrated Joule-Thomson derivative into existing Equations of State and compressible flow calculations.
  • Added JT Coefficient, slurry volume fractions for each component (4 component and Liu models) to results table, printouts, etc.
  • Increased accuracy of calculation of Control Valve Cv and %Opening for gas systems.
  • Added 2-phase loss correlations for control valves. Not covered by ISA guide. Used Parcol method.
  • Default Folders for License, Data, Preferences, and Templates moved to “C:\Users\Public\Documents\Flite\FluidFlow” for NEW installations.

FluidFlow v3.23

  • Added Flowsheet Undo facility.
  • Added ability to plot two-phase flow pattern charts for ALL pipe inclinations.
  • Added ability to plot composite charts for boosters in parallel and series. Access via New flowsheet toolbar button.
  • Added ability to plot HGL and EGL charts.
  • Updated Pipe Sizing Data.
  • DATABASE ADDITIONS – Over 50 new fluids added, 40+ Pumps and valves + many more fluid equipment items.
  • CHANGE: Density results shown for all phase states are now static density. Previously Stagnation Density was shown.

FluidFlow v3.22 Build 6

  • The calculated K value is now an available result (in export, excel, tables, etc.) for the following elements: GenericK, Inline Filters, Cyclones, Expansion Bends, Known Resistance Exits, Bends, & Mitre Bends.
  • Buried Pipe Calculation now uses an iterative solution process, which results in a more accurate solution. Log Mean temperature is calculated more accurately.
  • Now possible to “Check for Updates” from the Help menu.
  • NPSHa now shown for auto boosters and PD pumps.
  • Improved heat transfer in two-phase flow.
  • Heat Loss from Pipes – Instead of R values (heat transfer resistances) being displayed, individual transfer coefficients (Inside Film, Pipe Wall, Insulation, Outside Film) are now displayed. This change was requested by several users.

FluidFlow v3.32

With version 3.32, we have improved our physical property predictions and have introduced the ability to model petroleum fractions from ASTM D86 or True Boiling Point Curve data.

We have improved the accuracy of our two-phase flash calculations, added pressure recovery effects and generally improved the two-phase solution algorithms.

Our pre-release Quality Assurance tests have been expanded (we now test over 700 worked examples, covering all phase states).

For slurry calculations a new deposition calculation method “Oroskar and Turian” has been added.

The flowsheet results presentation has been improved and this is also reflected in the generated reports.

We continue to work hard on our upcoming Version 4 product which is due to be released later this year.


  • Added ability to model Petroleum Fractions to the Physical Property Estimator and the property database.
  • Added pressure recovery into 2-phase flash calculations. The effect of this becomes apparent with liquids at their boiling point.
  • Added new correlation for estimating Settling Slurry Critical/Deposition Velocity. Oroskar and Turian.
  • Added ability for a pump to handle low two-phase quality mixtures at suction.
  • Improved Two-Phase flow calculations by tracking enthalpy along pipe; this allows for more accurate flash calculations.
  • Added ability to show control valve charts from the flowsheet.
  • Improved flowsheet text and fly-by formatting.
  • Added ability to activate over a network.
  • Addition of an automatic backup to the DATA folder. Occurs every 30 days by default.
  • Added more QA examples. We now have over 700 examples that are fully checked before each new release: 85 Two-Phase, 176 Compressible, 79 Equipment Sizing, 251 Incompressible, 79 Non Newtonian, 11 Petroleum Fraction, 8 Saturated gas and Two-Phase and 22 script examples.


  • Velocities and Pressure for ALL size change elements are now based on the actual size of the element.
  • Added additional checks for out of range atmospheric pressure changes made by the user via the ‘Calculation Options’ dialog.
  • Removed ability to customize the Components Palette.

FluidFlow v3.30

  • Added NEW settling slurry calculation method. 4-Component model based on a 2007 paper by Sellgren and Wilson.
  • Added NEW settling slurry calculation method. Liu Dezhong method.
  • For settling slurries, size distribution charts can now be viewed at the Supply Node(s).
  • Added the ability to Size Relief Valves and Bursting Discs to API520 or ISO4126
  • Flowsheet – Added the ability to drag an OpenPipe node to make a connection to any other node (provided node can accept another pipe connection) OR to drag an OpenPipe into an existing pipe.
  • Flowsheet – Added the ability to rotate any node, in 90° increments for ortho mode and in 30° increments for iso mode.
  • Flowsheet – Added the ability to store a flowsheet as a template. Templates can be inserted into existing networks and connected via drag connect.
  • Added a resources window, that connects back to website for examples and videos.
  • Improved the temperature and pressure range and the accuracy of the internal relationships used to predict air physical properties.
  • Expanded the pipe heat loss calculation method, to enable direct entry of U values. Full calculation of U value is also available.
  • Added Specific Heat Capacity, Flow Cross Sectional Area and Out Flow Cross Sectional Area to results table, flowsheet properties, fly-bys etc.
  • Data added for PE100 Polyethylene Pipes, Non Newtonian fluids (phosphate clays, red muds, fly ash, sugar processing fluids and various foodstuffs).
  • Created a Licence Manager that also gets installed with the software, allowing greater licence flexibility.

FluidFlow v3.22 Build 5

  • Added ability to calculate heat loss in buried pipes.
  • Improved calculation consistency for non-Newtonian Casson and Hershel Bulkley fluids.
  • Input Inspector now highlights properties that have been edited (i.e., changed from default values).
  • All calculation modules are now available for each calculation, but user has a restricted view of results if a calculation module is missing.
  • Improved gas mixture physical property prediction. You can now use mole or volume fraction to define a mixture.
  • Improved gas calculations with heat transfer. Better integration of heat transfer equations into loss calculations and improve consistency of results.
  • New warning added to slurry module if user attempts to input a solids concentration greater than the packed bed voidage.

FluidFlow v3.22 Build 4

  • Added back the ability to decide on the Vsm calculation method for settling slurries.
  • Added results to text output.
  • Expanded scripting section of the help file.
  • Added some new PP and PPF Pipes and some new PP diaphragm valves into databases.
  • Stopped an “Out of Resources” error occurring in the scripting window if margins were shown.
  • Fixed a heat transfer bug that occured when 2 exchangers were used in series, with the Heat Transfer Option set to “heat transfer into network”.
  • Fixed a bug that occured if a dialog box was shown off screen.

FluidFlow v3.22 Build 3

  • Networks can now be imported and exported via text files.
  • Added ability to size pipes automatically – Beta Release.
  • Over 50 new fluids added bringing total of fluids in the database to over 1050.
  • Settling Slurry Calculations – Method extended to included inclined pipes.
  • Added ability to run help files locally from the network release.
  • Database mixtures can now have a fixed phase state, this is useful for users who do not have the 2 phase module.

FluidFlow v3.22 Build 2

  • The V3.22 upgrade now supports Paper/Pulp Stock calculations as a standard part of the Slurry and non-Newtonian module.
  • ASME and ISO 4126 calculation methods for predicting flow across safety Relief Valves are also available across all modules.
  • This release also provides the ability to export and import network designs via text files, meaning it is now possible to interface FluidFlow to other applications.
  • The dynamic analysis and scripting module has undergone a major internal reorganisation to speed improvements and additional functionality. Users can now write scripts in Basic as well as Pascal, for instance, users can easily call Excel directly from script with methods available to set and get Excel data and charts.
  • In response to user requests, the V3.22.2 release has improved UI speeds and additional abilities such as the specification of stagnation or static pressure at boundaries.

FluidFlow v3.22 Build 2

  • The V3.22 upgrade now supports Paper/Pulp Stock calculations as a standard part of the Slurry and non-Newtonian module.
  • ASME and ISO 4126 calculation methods for predicting flow across safety Relief Valves are also available across all modules.
  • This release also provides the ability to export and import network designs via text files, meaning it is now possible to interface FluidFlow to other applications.
  • The dynamic analysis and scripting module has undergone a major internal reorganisation to speed improvements and additional functionality. Users can now write scripts in Basic as well as Pascal, for instance, users can easily call Excel directly from script with methods available to set and get Excel data and charts.
  • In response to user requests, the V3.22.2 release has improved UI speeds and additional abilities such as the specification of stagnation or static pressure at boundaries.