Drum Level Measurement


Thread Starter


In investigating some anomalies in my HP drum (900-1800 PSIG) DP level measurement, I have found that my transmitters are ranged higher than the mechanical measurements. The application is a wet leg measurement, drum face tap penetration, drum bottom penetration, heat trace, condensate pots.

All three of like transmitters are ranged the same on three different sets of identical taps on opposite sides of drum. xmitter range 88 in wc, tap distance 54". Drum control set point should be -22 from top tap +/- 3"

I have similar issues on all three drums, of course they are all different ranges.

I can not figure out why these transmitters are ranged this way?
Steam Drum levels are quite tricky. I know just enough to leave it to the experts.

I'm sure there is a good explanation.
The transmitters should be ranged for the differential pressures they will measure at the operating pressure and at saturation conditions. There is a significant difference in the water and steam densities at operating condition as compared to the cold condition. Do some research, this subject has been covered on Control.com.

As CSA often says, use the handy search box in the upper right corner of this web site. And don't change the transmitter ranges until you fully understand the process!

Hilton Lawson

The density of water changes with temperature and pressure. Unless you are experienced in this calibration I would recommend that get help before tackling it. If you doubt the current calibration, there is a good article on Emerson.com on this application. You should read and understand the article before proceeding.
Don't change settings and calibration without the design specs, mech drawings, mechanical, and operations getting involved.

With cold startup, you may need to know the level over a larger span,
with both the condensate and steam density taken into account as the system is brought up to operating conditions.

The drum level calculations will be part of the engineering documentation of the boiler. If not, then you need to get the boiler designer/supplier involved. Boilers are complicated systems with a lot of safety issues.

The 88" span does sound curious and may refer to boiler swell that can occur during extreme demand or loss of firing, rather than the transmitter span.


The density of water/vapour mixture in steam drum at operating conditions is typically less than that at atmospheric conditions. This means that the calibration span of transmitter is typically shorter than mechanical measurements.

You stating that the transmitter range is higher than the mechanical measurements is surprising, because it is normally the other way around.

By the way, we all assumed that we are talking about steam drums, although you didn't mention anything in your question. Can you confirm?
Yes it is HP steam drum the instrument is spanned 0 to -82 "wc and the tap measurement is 54". I do understand the nature of compensating for the density and it makes math sense when calculating. I just needed to confirm this fact because I have in internal (DCS) algorithm that is also compensating based on drum pressure. I think this is redundant and not necessary.


If you are compensating in DCS using drum pressure, there should be no density compensation in the transmitter span. They should be calibrated so that the span is same as the physical tapping distance.

If you do not want to use DCS compensation, you need to remove the function block doing the compensation calculations and incorporate the density factor in span calibration of the transmitter. However, the drum pressure compensation is usually a preferred method, as it gives you correct reading at all pressures, not only the operating pressure, which is helpful in startup and shutdown of the boiler.