A strainer is a device installed in a piping system which provides a means for mechanically removing undesired particles from a flowing fluid.  Most strained particles are in the size range between 40 micron and 1 inch, and are typically removed by using a perforated, mesh, or wedge wire straining element.  The purpose of straining particulates is to protect downstream mechanical equipment such as pumps, heat exchangers, control valves, flow meters and spray nozzles from the detrimental effects of flow debris.  It also serves to prevent this debris from ending up in the final product in some manufacturing cases.

There are two common types of strainers, the “Y” strainer and the basket strainer.  The “Y” strainer has one fairly universal design and can be used in horizontal and vertical piping applications as well as high pressure applications.  There are many varieties of basket strainers, including simplex (one basket), duplex (two baskets), and automatic self-cleaning strainers.  All operate on the principle that the fluid flows through the straining element which captures the debris.  At steady state conditions, the pressure drop is a function of the square of the flow rate: 

dP = (Q/Cv)^2. 

As the straining element fills up or becomes plugged, the pressure drop through the strainer rises until it reaches some preset value.  At that point, the strainer element is removed and cleaned, or the device is switched to the other clean element.  Some devices have automatic backwash or blow-down cycles so the operator does not have to perform the cleaning cycle.

The fact that the straining element gradually becomes fouled or plugged over time, presents a unique situation when attempting to model a strainer in PIPE-FLO.  A strainer is a variable resistance device, and therefore the pressure drop across the strainer will vary not only by the rate of flow going through it, but also by the amount of debris which has been collected.  A strainer could be represented by a flow coefficient (Cv value), but as the strainer becomes fouled, this Cv value would have to become smaller and smaller.  Similarly, a strainer could be represented by a resistance coefficient (K value), but as the strainer becomes fouled, this K value would have to become larger and larger.

One way to solve this problem of a time-based variable resistance is to just model the strainer’s worst-case scenario for pressure drop.  This would be a conservative approach, and would ensure that the actual strainer outlet pressure would not fall below that which was being calculated.  The way to do this would be to model the strainer as a component, and set this component to a fixed dP value equivalent to the preset limit on the strainer.  The figure below shows a system with a strainer just upstream from the pump.  This strainer is set to a fixed dP of 10 psid.  This means that the pressure drop will be 10 psid across the strainer regardless of the flow rate.



One piece of noteworthy information that this example shows us is that the net positive suction head available (NPSHa) to the pump when the strainer basket is full is only 18.85 ft.  In some pump applications, this might be on the verge of causing cavitation.  In any case, modeling a strainer with a fixed dP set to the preset limit will give you the highest expected pressure drop through the device.  Normal operating conditions would produce a lower pressure drop.

Important Note:
Strainers are one of the few devices which should ever be represented as a fixed dP component.  Heat exchangers, chillers, air handling units and most other devices should never be modeled as fixed dP components because as the flow through the device changes, the pressure drop changes.  If these devices are modeled as fixed dP components, then any change in the system which varies the flow rates will result in inaccurate calculations.

You must also be careful about how you are using fixed dP components within PIPE-FLO.  If used incorrectly, the program may encounter convergence issues.  This can occur when there is unconstrained flow through a device modeled as a fixed dP component.  Unconstrained flow means that the flow through the device is not specified by other system devices, and instead must be calculated.  During the iteration process, a fixed dP component cannot be “balanced” like other devices which have a flow vs. dP relationship.  In some cases, especially if there are multiple fixed dP devices, this will cause the program to not converge on a solution within the set tolerances, and you will see the following message:



If this happens, then the results will likely not be as accurate as a system which has converged completely.  In this case, you should find a way to specify the flow rates such that the flow rate through the fixed dP device does not have to be iterated.  Or, the other option would be to use curve data for the device rather than a fixed dP.