Today is...
Wednesday, February 21, 2018
Welcome to, the global online
community of automation professionals.
Featured Video...
Featured Video
A demonstration of EtherCAT control of linear motors using the CTC EtherCAT master.
Our Advertisers
Help keep our servers running...
Patronize our advertisers!
Visit our Post Archive
Three Element Drum Level Control Problem
Drum level unstable at lower turbine load
By Rahul P Sharma on 9 January, 2008 - 1:32 am

We have a waste heat recovery boiler that is supplied by exhaust of a 20MW Gas Turbine. We've seen that at lower turbine loads (75% and below) the three element drum level controller cannot maintain the drum level at desired setpoint. As soon as the load on the Gas Turbine is increased to more than 75% of rated load, the stability keeps getting better. At rated load (20MW) the drum level is very stable and close to the setpoint.

Can anyone please help with information on how to keep the drum level stable at lower turbine loads?

Does your WHB have supplementary firing or is it only Turbine Exhaust Gas? What are the Steam conditions (Pressure, Temperature & Flow).

If your steam demand is relatively stable then you only really need single element control (Level) 3 element control is used to reduce variations as a result of changes to the Feedwater flow and steam flow. In theory if the water into the boiler is equal to the steam flow out then your drum level will be stable.

However as your Gas Turbine load drops <75% the heat input to the boiler reduces causing the Steam Drum pressure to drop, as the drum pressure drops the level will rise, this will effect the Steam flow as I assume your measured steam flow is corrected for Temperature and Pressure. As your measured steam flow appears to increase due to pressure drop the controller will attempt to maintain the feedwater flow to match the steam flow, at this point your control system is very unstable.

If you need more info send your e mail.

Every control loop (and that includes the controller, measuring instruments, and controlling element) will have one set of tuning parameters optimal for one operating regime only. All the other operating regimes will then have to be a compromise between stable operation (i.e. no offshoots) and sluggish operation (slow reaction to a change). Stable operation does not mean that the actual measured parameter (in your case the water level) remains at one point only, but varies within a band such that the band's mean is the setpoint you are seeking to achieve, without the upper and lower limits exceeding allowable levels. All control systems will finally settle to the setpoint, as long as no external changes to the process are made. It is just a question of time.

In your case it seems that the water level control loop is tuned to be in perfect control at around 100% gas turbine load. To achieve better controllability at lower loads you have two options:

1. Re-tune your controller such that the optimal control point is achieved at a lower gas turbine load, e.g. 75%. In this case, then your process oscillations will be lower at the lower loads, but you may experience some sluggishness at loads close to 100%.

2. Replace your controller with one that gives the facility of changing control parameters with signal output level, i.e. when signal to control valve is 50% or less (for example) it uses one set of PID tuning parameters, and at signal levels higher that 50% it uses another set. In this way you can have two optimally tuned operating regimes, and the other operating regimes will then operate pretty stable.

Obviously all the above depends on:

1. How you operate your gas turbine/HRSG, especially how the combined unit is ramped up and down.

2. The size/capability of your feedwater control valve. For optimal stable operation, your control valve should be 75% to 80% open when the gas turbine is running at 100% load under steady state conditions. If your valve aperture is less than this, most probably you have an oversized valve for your application. This will practically mandate that you change your controller to one that uses at least two PID tuning parameters (described in 2 above) as it would be practically impossible to tune it for the whole operating range of your unit.

I hope this gives you more insight to your problem.

By Rahul P Sharma on 11 January, 2008 - 9:05 pm

Thanks sandy, jojo for your inputs. Here are some basic inputs of our system.... I should have included them in my first post itself.

Its a 90TPH boiler that runs mostly in GT Exhaust Mode and occassionally on both GT Exhaust and Supplementry Firing mode.

With GT Exhaust alone the HRSG supplies 60TPH of fairly constant plant steam demand... The controller is configured in a Honeywell's TPS/TDC3000 DCS system. The operating boiler pressure is 10.5 Kgf/Cm2. Steam temp is 180 DegC.

The valve opening at rated GT load is around 55-60%.

We have an auto switchover facility that puts the controller back to single element mode for Steam load less than 30TPH. There is a separate control valve for Single Element control... The three element valve closes after the controller switches to single element mode... But this happens very rarely and is not a problem.

The problem we face is only for GT loads < 75% of rated 20MW.

Sandy, you mentioned that you have some more literature on this... Can you please mail it at rahulpsharmaATgmailDOTcom, if you can... or even suggest a specific link for further reference?

au revoir

At lower loads, the flow of heated gas may be stratified. This would cause uneven heating of the heating surfaces of the boiler. The drum level transmitter might be detecting these variants in actual drum level. Watch drum level transmitter output closely, if this is the case, try to suppress the fluctuation with PID changes on the water level controller.

Also a three element system can cause water level offsets if the feedwater and steam flow transmitter do not mathmatically match. Check the calibration and range/span settings of both transmitters to ensure that water and steam match pound for pound in regards to flow throughout the entire range.

Last but not least, the entire system was probably tuned (PID) for maximum load.

You may have to tune the controllers to another load to get acceptable performance at both low and high turbine output.