Boiler Drum Level

R

Thread Starter

Ritika

I have a question to the experts of this forum:

1) If the feedwater supply pressure is fairly constant, then is it possible to eliminate the third element - feedwater flow sensor?

2) I want to get some numbers about three element control. Say,
FT Steam = FTS;
FT Water = FTW;
LT Drum = LT

Request to kindly help me understand the interconnections wrt to a 3-element Drum Loop.

I understand that 4-20mA of LT gets connected to the LIC_Drum. The difference of the Drum SP of say, 50% (12mA) gets subtracted from instantaneous (say 8mA/25%) 4-20mA from LT. This error (4mA/25%) gets acted upon by the PI/PID algorithm of Drum LIC. The LIC_Drum output is a 4-20mA PID output, say, 75% (16mA) (assuming a direct action controller output).

THIS IS WHERE MY QUERY COMES IN.

After this step the FTS's (FT_Steam) 4-20mA should get added to this LIC Output as a Feedforward path, right? If, say, FTS is 12mA (50% steam flow), then will this 12mA get DIRECTLY ADDED to the LIC_Drum's mA PID Output? i.e., Will LIC_Drum output of 16mA get added to 12mA of FTS? But this doesnt make sense since the summation is 28mA now? Do we need to scale the FTS's mA? If so then by how much? Will the summation of LIC_Drum Output & FTS never cross 20mA? Should the Feedforward Path Gain be chosen to scale FTS to keep this summation below 20mA always under all conditions? I am really lost on this. Please help me understand, what is it that I am missing.

Thanks for your time
Regards
Ritika
 
B

Bruce Durdle

There are several different approaches. The one I favour, because it is easy to understand, is to say that you want the feedwater flow to match the steam flow at all times. So the feedwater setpoint needs to track the steam flow setpoint without scaling:

FIC0101_SP = FT0112_OP where F0101 is the feedwater flow tag and F0102 the steam flow tag.

Now you need to add in a corrective signal or bias from the level controller output, to adjust for measurement errors and blowdown, etc. So:

FIC0101_SP = FT0112_OP + LIC0105_OP.

As you say, this will potentially go out of range. To avoid this, add in a second fixed bias term of 50%:

FIC0101_SP = FT0112_OP + LIC0105_OP - 50 %.

The level controller output will be at 50% +/- trim when the level is at setpoint and can go either way in the event of an upset, without going outside limits.

Bruce
 
C

Chandrakant Prajapati

Dear Bruce,

In the Equation

FIC0101_SP = FT0112_OP + LIC0105_OP - 50 %.

You have asked to subtract 50 % from the sum of Controlled output of Level PID and Pressure & Temp Compensated Steam Flow. This 50% remain Fixed or it to be dependent on any variable?

Please let me know your answer on my mail ID i.e. chandra_n [at] vsnl.net
 
B

Bruce Durdle

Hi Chandrakant,

The 50% is simply a bias to allow the level controller output to increase or decrease the steam flow signal - without it, the level controller output will be at 0% when steam flow and feed flow are equal. There is no reason why it cannot have any other value if for example the system is affected by disturbances that make the level go much higher than the low extreme.

Cheers,
Bruce
 
Top