2021 AFPM Summit Virtual Edition: Improving plant operations with control valve simulations
REID YOUNGDAHL and LUKE NOVAK, Emerson Automation Solutions
Valve testing ensures that valves meet the requirements of an application—as far as plant personnel can estimate. However, in some cases, conditions are not exactly as predicted, and the valve may not perform as expected.
In such cases, plant personnel are faced with an expensive decision: Should the valve be removed and sent back to a lab for further testing, or is there another solution?
With computational fluid dynamics (CFD), a simulation engineer can develop predictive, physics-based computational models to cover a wide range of tests and situations, including valve flow coefficients, multi-phase applications, turbine bypass temperature sensor optimization, installed valve troubleshooting, and more. With CFD, a simulation engineer can input the design of the valve in question, simulate the installed conditions, identify the problem and develop a solution.
CFD at work. CFD simulation is already being used by most valve vendors for initial design (e.g., when engineers must understand a valve’s expected flow performance under certain conditions). An experienced valve designer typically has a sense for how the valve and trim geometries must be shaped to provide the desired flow performance, but CFD provides insights and physics-based predictions of the expected performance.
Physical flow testing has historically been used to determine valve performance, but CFD simulations are now commonly used to predict flow performance during the design process of new valves, either prior to, or in lieu of, physical flow testing.
Typically, a valve designer will create a valve, use simulation to evaluate concepts and refine the design (FIG. 1), and then have a valve cast. The new valve will then be subjected to flow lab testing. If the valve performs as expected, it can go into production. If not, testing will usually point out areas requiring modification. The valve design simulation model can also be used for future testing, including CFD simulation of valves installed in the field.
FIG. 1. This diagram depicts a fluid flow distribution through a valve as predicted with CFD.
The importance of validating CFD simulations against actual test results in a flow lab (FIG. 2), and the involvement of analysts with suitable competency in simulation to assist in ensuring the quality of results, is crucial.
FIG. 2. This Emerson Innovation Center flow lab in Marshalltown, Iowa, verifies the results of simulations by testing the valve under expected operating conditions. Once the simulation model has been verified, it can be used to solve problems reported from the field.
Simulation experts apply best practices with quantified uncertainties developed using data procured in test facilities, coupled with field experience, to provide value in ways only simulation can provide. End user collaboration is critical to identifying the pain points and understanding where simulation can be of service.
Some simulation capabilities include:
- Fluid dynamics analysis via simulation of capacity, choking, velocities and pressure profiles
- Validation through prototype and production unit flow test
- Structural checks using finite element analysis
- Validation through digital image correlation and strain gauge hydro tests
- Thermal analysis via computational model thermal profiles
- Validation through prototype and production unit process temperature tests
- Seismic analysis to computationally predict and exaggerate structural loads
- Validation through prototype and production unit load tests.
Traditionally, CFD has only been used as an internal resource by valve vendors—primarily for new product development and testing—and is rarely made available outside a vendor’s lab. However, some control valve vendors are now providing CFD simulation services to end users to help them cut costs, reduce downtime and make other operational improvements, as with these examples.
CFD replaces testing. A major OEM of power plant turbines required flow capacity Cv validation on a 32-in. Fisher high-performance butterfly control valve installed in a critical bypass application.
The traditional solution would have been to physically flow test the valve, but this was deemed unacceptable due to increased costs and negative schedule impact.
Instead, Fisher valve simulation engineers created a CFD model of the butterfly valve to simulate the Cv, then verified to the OEM through a report that the valve would perform as expected. The simulation provided a ±7% confidence band for all flow coefficients.
The turbine OEM saved more than $100,000, with no delivery schedule interruption, by using CFD as an acceptable alternate to physical flow testing.
Accommodating installation requirements. Thirty years ago, a refinery installed a Fisher Type 461 severe service control valve in the non-preferred flow-up direction. The refinery is now undergoing an expansion project where process conditions require a larger capacity, severe service control valve. The refinery wanted to replace the valve with another Fisher Type 461 due to the 30+ years of proven service, but they realized that changing the piping to the preferred flow-down orientation would be too costly and negatively impact the project schedule. The refinery wanted to keep the orientation flow-up, which led to sizing implications since the valve sizing parameters were only flow tested in the flow-down orientation.
The service was too critical to size the valve using the process of extrapolating valve sizing parameters to match the application. Instead, a CFD simulation was performed to predict valve flow coefficients and pressure drop ratio factors, allowing for accurate control valve sizing. The new valve was installed and is functioning as required in this severe service application.
Put the temperature sensors where? CFD provides value in both front end engineering design (FEED) and post-installation situations. This CFD application solved a problem and provided a model for this end user’s future FEED projects.
Turbine bypass is considered one of the most critical control valve applications in a power plant. Properly selected turbine bypass valves are important for keeping a turbine safe and maintaining overall power plant heat rate, a common measure of efficiency. In a bypass system, steam is de-superheated by creating a pressure drop, with a control valve used to add the proper amount of water to the steam.
CFD was used to predict the ideal location to install temperature transmitters, specifically where the added water is fully evaporated, and where temperature readings would therefore be most accurate. With requirements for turbine bypass valve temperature transmitter locations better understood, downstream straight-length piping requirements were reduced from those provided by the sizing tool. This cut installation cost and time.
An undersized actuator. A high-pressure injection pump recycle valve on an offshore platform in the North Sea was experiencing instability issues when trying to control a 3,700-psid seawater pressure drop. The solution for sizing this critical severe service, multi-phase, fluid control valve was using CFD to ensure the control valve and the downstream pressure relief valve were sized correctly. For both valves, CFD simulations were used to predict valve Cv and pressure drop ratio factors, allowing for accurate sizing.
In addition, root cause analysis of an undersized actuator on the control valve was derived using both smart positioner online diagnostics and CFD simulation. This was accomplished without taking the valve apart and causing further interruptions. A new actuator was shipped and installed, solving the problem at minimal cost and disruption to existing operations.
Outside industry standards. Control valve sizing standard IEC 60534-2-1 covers most control valve applications. However, there are still many applications that fall outside its scope, such as outgassing and fluid flows composed of multiple phases.
A refinery water treatment skid OEM required CFD analysis on its severe service, multi-component, fluid control valve application to ensure the control valve was sized correctly. The OEM didn’t want an oversized valve because this would negatively affect the sizing of the downstream pressure relief valve. The high pressure drop and presence of solids required a flow-down, severe service control valve with erosion resistant materials, including solid stellite trim with ceramic inserts, and an outlet liner extending beyond the valve outlet.
OEM process engineers provided full fluid composition data, enabling the valve simulation engineers to generate a CFD report and correctly size the control valve. Accurate CFD modeling mitigated the OEM’s risk of under or over sizing these final control elements, and instilled confidence in the OEM’s customer regarding performance of due diligence for these critical skid components.
Takeaway. When a valve vendor designs a valve using CFD—and tests the designs in a flow lab—it produces a software model that can be used to predict or diagnose problems with installed valves. CFD simulations are now being used in not only new product development, but are also being provided by valve vendors as a service to end users, saving time and money when diagnosing problems with installed valves.
ABOUT THE AUTHORS
Reid Youngdahl is a Valve Technical Specialist at Emerson Automation Solutions in Marshalltown, Iowa. He began working for Emerson in 2013 as an Applications Engineer, providing global control valve technical support for various industrial markets such as oil and gas, power generation and original equipment manufacturers. Mr. Youngdahl provides technical consultation for severe service control valve applications to optimize customer solutions.
Luke Novak earned a BS degree in aerospace engineering from Iowa State University, and has 12 years of industrial experience using CFD. Since joining Emerson in 2017 as CFD Lead in Marshalltown, Iowa, his focus has been CFD analysis of process control valves. Mr. Novak’s responsibilities include applying CFD to solve business critical problems, while developing best practices for deploying CFD within the wider engineering community.
Comments