"Ask the Experts" is moderated by Béla Lipták, process control consultant and editor of the Instrument Engineer's Handbook (IEH). Preparation of the 4th edition of Volume 3 "Process Software and Networks" is in progress. If you are qualified to update an existing or to prepare a new chapter, please let me know. Similarly, if you feel qualified to answer questions in this column or want to ask a question, write to me at: firstname.lastname@example.org.
Q: I'm looking for some information relative to a controls engineering project I'm working on. I thought you might be able to point me in the right direction. We manufacture large gas-fueled internal combustion engines. Many are designed to run on various types of gases, such as natural gas, propane, digester, landfill gas, etc. In our engineering lab we have a need to mix natural gas with propane or CO2 gases to simulate engine operation in the field across the spectrum of field gases encountered. The project involves coming up with a system to mix two separate streams of these gases continuously across a wide range of mix ratios, with automated ratio control and with automated capability to handle mix demand changes due to decreased/increased demand from the engine fed from the mixer. Precise control of the mix ratio during the transient of a demand change may not be necessary, but there likely needs to be some provision to increase the flow of each stream so the ratio doesn't stray too far. A simple controller could then be used to fine-tune the mix ratio to the desired setpoint once the new demand has been satisfied. I'm envisioning a skid with ordinary gas regulator(s) to handle the demand change, and perhaps a control valve(s) for fine-tuning of the mix ratio. Pressures are low, 60 psig max. We generally look to hold the mix output to a relatively constant outlet pressure. That constant outlet pressure then feeds an engine-mounted gas regulator. We would install one of these systems in every engine test cell. I'm looking more for some guidance as to the appropriate configuration of such a system to provide the functionality required, preferably using standard gas regulator(s), control valve(s), a simple PID controller, etc. These would be lab-only systems. Any thoughts on where to go for reference material and/or subject matter experts for help with this project?
A: Below you will find the configuration of a blending system (Figure 1). The power output of the engine can be modulated by varying the total flow setpoint of FIC-1.
The fuel blend ratio of the gases can by adjusted by varying the setpoint of the ratio setting of FY. The setpoints can also be adjusted manually or automatically. If automatic blend composition control is needed, you need to measure the blend composition and arrange the composition controller to act as the cascade master which is modulating the set point to FY.
If the plant is controlled by microprocessor-based digital control or DCS systems, the configuration of such control loops requires much less effort. If the flow measurement signals are transmitted utilizing bus-based technologies, the associated dead time should be considered in the tuning of the flow ratio control loops because their time constants are short.
A: This looks like a fine application for control in the field, as provided by Fieldbus Foundation devices and simple laptop applications to configure and control the devices. There's no need for a DCS. But you didn't specify the maximum flow rate and the number of cells to be controlled. The time constant of the valve determines the loop sample rate, which depends on the size of the actuator. Ten times a second is about the maximum sample rate for fieldbus devices. That matches a really small and fast valve. You did say large engines, so maybe not so small. Do you have instrument air for the valves?
You didn't specify the accuracy required, so perhaps something cheaper than fieldbus devices could be used. Digital devices will simplify logging of parameters, though they can cost more. Alarms for out-of-tolerance conditions do not cost extra with fieldbus devices. Do you need mass or volume flow measurement?
The number of loops required has a cost cross-over point between single-loop controllers and a small fieldbus system. Fieldbus Foundation control information can be found Fieldbus Foundation control information can be found beginning at www.fieldbus.org. Ian Verhappen at www.industrialautomationnetworks.com and John Rezabek are founts of non-marketing information. (Editor's Note: John's On the Bus column "Surprise! Field-Based Control Beats DCS".)
My website has a manual for control in the field at http://www.hlq-ltd.com/
Q: I have a question about sizing for the body of a control valve. Is it mandatory that pipe line schedule be mentioned for control valve sizing? If ye,s then what is the role of pipe schedule in control valve sizing?
A: I would stress the fact that the valve rating should be the same as the pipe, independent from valve size. There could be a contractual constraint calling for minimum rating 300# for valves up to a certain size (3" or 4", according to the customer specifications).
Pipe schedule can play a role in determining the velocity in the outlet piping that has some constraints (sometimes contractual or good engineering practice). If the schedule is very high (special thickness for very high design pressure and temperature), the area of the pipe is much smaller than with a lower schedule, with consequent higher velocity. Sometimes, it is not possible to specify a body size equal to the trim size because of too high velocity inside the valve that could produce cavitation, flashing or erosion, thus reducing valve lifetime.
In case of valves with butt welding ends (typical of power industry) the piping and the valve internal and external diameters should match exactly. This is the case where the piping schedule is essential for valve manufacturer.
A: I understand that the question concerns the line schedule rather than the flange rating. In most cases the pipe schedule probably won't matter much as the impact on valve sizing will be minimal. However if there is cavitation and/or flashing, then the pipe schedule will impact the approach and exit velocities, and it can have a significant impact on the sizing calculations. It is always better to include the information if possible.
A: A DN25 (1NS) control valve can have a valve coefficient of Cv=12 or even higher. Such valve can be installed in DN25 Sch 40 pipe, or even Sch 80. BUT many piping systems specify Sch160 or XXH for DN25 for mechanical strength or your DN25 valve may be in DN50 pipe, chosen for support strength. In such a case, the bore of the connecting pipe can be limiting the Cv to a maximum value of Cv=6. In such a situation, I would avoid using DN25 valves, and use DN40 because the price is not much higher, and the chance of embarrassment much lower. The swage correction practice using the nominal pipe size just doesn't work at the bottom end of the valve size range.
I recently saw a small valve installed which was trying to pass Mach 1.5 through the swage. It failed to do so. The pipe schedule is a significant component in the noise calculation – Sch80 pipe can give 6dB reduction relative to Sch40. So the valve supplier needs to know the downstream pipe schedule (and material) as well as the size. In gas flow, the guideline is to keep the valve body exit velocity below 0.3mach (absolute maximum 0.5mach if using a close coupled noise-reduction restrictor.
Ian H. Gibson CPChem RPEQ FSEng
A: The concern is around the stresses caused by thermal variation and imperfect support which a large pipe can exert on the valve body. One simplistic rule of thumb is a avoid valve bodies sizes that are half or less of the nominal pipe size. This is far too general.
I classify control valve designs as heavy duty, medium duty and light duty. For valves which dissipate high energy, the pipe wall thickness has a significant effect on aerodynamic noise radiation through the pipe. For severe cavitation or severe corrosion, a heavier pipe will stand up for a longer time, but the body material may also make a difference.
As I see it, the pipe schedule has very little effect on the flow capacity of a control valve except in extreme cases, and it is necessary to match the flange classification used.
Cullen Langford PE
A: The only time the downstream pipe schedule should be considered in sizing of the valve is in predicting the noise level. If the noise level is too high, valve design changes can be made to bring it down. This may only mean the selection of a different valve trim. If the primary noise source is the outlet jet from the valve, increasing the valve size will reduce the noise.
The in and outlet pipe schedule will also impact the weld prep shape for a valve that meets ASME weld prep standards. However, I would not list this as a sizing issue.
Herb Miller, PE