Comparing steam flow and feedwater flow


Thread Starter


I know that in boiler control, flow rates for steam and feedwater cannot be compared directly as they have different densities.

What I'm looking to find out is how those values (ones that might come from a flowmeter) are compared in order to have a mass flow rate comparison?

How must these values be manipulated?

How are pressure/temperature compensations made?

Any help would be greatly appreciated.

Bruce Durdle

I don't know why you feel that the flow rates cannot be compared directly as they have different densities. You are not comparing the flows directly; you are comparing the results of the flow measurements, and the flow transmitters can certainly be set up so the outputs can be compared directly. The configuration or hardware design will incorporate an adder as comparator block, configured so that the steam flow signal is subtracted from the water flow signal.

In a 4-20 mA system, it is relatively important to make sure that the steam and feedwater flow transmitters are ranged the same. If this is not done, the comparator block has to be set up to incorporate a suitable scaling factor - eg, if the water flow range is 0-120 tph and steam flow range is 0-100 tph, you need to get them both into one or other range - possibly by using a 1.2 x scale factor on the water flow.

In a DCS using engineering units, this is all taken care of and there is no need to do anything in particular.

Pressure and temperature corrections will usually not be applied to feedwater, but are commonly used for steam. For these you need a pressure and temperature reading, and a suitable calculation block or module. Details will depend on the accuracy you require, but this is not usually too challenging.
As long as you compare mass flows they can be compared. So you have to compensate the steam flow measured differential pressure with steam pressure and temperature to get mass flow, and you probably should compensate the feedwater measured differential pressure with temperature to get mass flow.
What I'm doing is a research project that entails changing a single-element to three-element control.

I have two identical vortex flowmeters, both providing a 4-20. 1 is on the steam line and 1 on the feedwater line. What I'm unclear on is how those signals compare in order to maintain mass balance in a boiler. Ill be detailing how this would all interact with a PLC system to control it.

Thank you for the help.

Bruce Durdle

You're not going to be able to maintain a perfect mass balance - for 2 reasons.

First, the feed flow measures the total feedwater into the boiler - but the steam flow meter measures only the water converted to steam. You won't get a measurement of the blowdown used to remove unwanted chemicals from the drum.

Second, even with two identical flowmeters there will be some error - hopefully less than 0.5% but any very small error will be integrated by the action of the drum - so if you try to maintain the feed flow exactly equal to the steam flow, and have no blowdown, the drum level will rise or fall at a rate depending on the error.

In the simplest real-world implementation, the drum controller o/p and steam flow elements are combined in a calculation block having the equation

Out (%) = Drum controller o/p (%)+ Steam flow (%) - 50 %.

This signal is then fed to the SP input of the feedwater controller.
Since the drum acts as an integrating element, you don't need an I term in the controller - but one is usual to keep the operators happy. I have found that a setting equal to at lest 1.5 times the drum holdup time is acceptable.
research project?

Gee, this sort of change has been routine for the last thirty years, where the boiler service requires it.

What is you boiler capacity and what is the nature of the steam demand on a percentage basis? You also need to have performance data from you BFW suppply.

The steam drum control can be tricky and may require far more instrumentation than you currently use.