Desuperheater control in variable pressure header.

[rmw]

I need a control scheme for a temperature controller that will follow header pressure, and vary the desuperheated steam temperature set point according to a set margin above saturation pressure for the actual header pressure. The header is mechanical drive turbine exhaust, which typically has up to 50F superheat in it, and it is used for heating steam in tubular exchangers. The header pressure can vary from 5 psig to 20 psig, and do it often. Are there any controllers that can utilize a "look up" function to corelate pressure to saturation temperature, and thereby change the set point to follow the header pressure. As can be seen, by setting a single control set point for the highest pressure, 20 psig, then at the lowest pressure, there will still be 30F of excess superheat still in the steam, almost as bad as before we desuperheated.

rmw

 

[davefitz]

You should be able to input into the DCS a simple plot or linear equation of sat temp vs pressure for the small pressure range 5 psig to 20 psig. Its done all the time.

The setpoint for the temperature controller must be set at a value of superheat that is greater than saturation by at least the measurement accuracy of thermocouples. For standard uncalibrated thermocouples this may be 6-8F.

Typical attemporator logic that is based only on temperature feedback uses a 20 F superheat as setpoint. Also, the thermocouples must be located far enough downstream that they are not wetted by the spray , usually a distance of 0.3 sec transit time is the ideal downstream location for the thermocouple.

Another recommended addittion to prevent waterhammer in downsteream piping ( due to poor turndown of temperature controllers) is to include an override to prevent the ratio of spray water flow : steam flow from exceeding a limit of about 0.25:1.

 

[rmw]

I forgot to mention that this would not be a plant with a "DCS" level of technology sophistication, so it will have to be a stand alone, single loop type controller.

I agree with your comments regarding temperature control, but I have some prior experience in this area, and know some "tricks" that will get me closer than the "textbook" standard that I recognize must be imposed on those who won't pay the proper attention to detail, or who don't really know their process. The spray ratio control is just one such "trick" that can be used to good effect, and I have used it before.

These will be steam assist type desuperheaters, with high temperature condensate for the desuperheating medium, so I can get the fine atomization needed for tight control.

And, in similar type manufacturing plants, where the turbine steam from the boiler is generated saturated with no superheat at all, naturally, the turbines produce their exhaust wet, so the identical HX equipment in one of those plants might be seeing exhaust steam with as much as 13% moisture.

So, I am not afraid of some moisture, as long as I know I have the Spray : Steam raio you mentioned. The Hx's are pretty steady state sinks for the steam. Their reynolds numbers are very low, so erosion from the moisture is not so much the problem, but with such low reynolds numbers, a little superheat just "kills" them.

I plan to operate closer than most would, hence the reason I want to follow the satuation line with respect to pressure.

Thanks for your reply. I still need ideas.

rmw

 

[davefitz]

If you wish to operate down to the saturation enthalpy, or even to a controllable wet steam region, then just use the online enthalpy balance technique. Both CCI and Fisher Contek provide control algorithms and probably a programmable controller which will allow you to achieve this .

As long as the inlet steam has a knowable enthalpy( ie, initially superheated and can montitor the Pi and Ti), yu can compute the inlet enthalpy and determine the required spray water flow to reach an outelt enthalpy of any lower value. You will need an indication of steam flow and water flow, of course.

 

[rmw]

OK, I am digesting this. I don't know if this client has adequate steam flow measuring devices, (we have 4 individual points to desuperheat) or not. I am sure they have, or can put, Ti, and Pi, easily enough. I will check out the direction you have given. Thanks for the help.

rmw

 

[platte]

We have an absorption-style desuperheater from Schutte and Koerting. It is very difficult to control because of radical behavior of the temperature elements controlling the steam condensate input. Sometimes, no matter how much condensate is being added, the temperature does not change. This is more prevalent at low flow rates. Any ideas?