6 replies to this topic

[JLMONTREAL]

 Cooler_Temperature_control.xls   18KB   735 downloadsDear all, please give your idea.
There is a temperature control loop across a cooler as attached. (I appreciate Art’s EXCEL file which help me to finish the attached draft quickly.)

The hot fluid A at 150C is cooled by a cooler. There is a bypass control valve helping to maintain the target temperature at 50C, particularly in winter. I believe this loop will function well if the fluid is passing continuously.

However, the ZCV valve will open (once per 10 hours) and close (10 hours after being opened), then wait for another 10hours to open again. To my understanding, the temperature control loop won’t achieve steady target temperature (50C) in the first certain minutes when ZCV is open after 10 hours waiting if the loop is set on AUTO mode control.

The TIC indication will be first much higher than 50C (TV may be fully open during waiting period), even may be closed to source temperature 150C and then lower than 50C, finally will reach 50C after cycles of fluctuation. This is not appropriate in terms of process and valve maintenance requirement, particularly for large size valves.

So, if my understanding is correct, manually operating the TV in the first several minutes may be better to get stable target temperature. But the designer wants to keep auto control.

To avoid the valve shaking severely or temperature fluctuation, is there a better control strategy to automatically maintain steady temperature (say 50C) at TIC during all the time for this kind of intermittent operation?

Thanks.

[Andrei]

I have some questions:
What is the hot fluid?
Why you cannot put the control valve on the cooling water circuit.
As I know the response speed of the valve depends on the tuning parameters of the controller. You cannot tune the controller?

[JLMONTREAL]

I have some questions:
What is the hot fluid?
Why you cannot put the control valve on the cooling water circuit.
As I know the response speed of the valve depends on the tuning parameters of the controller. You cannot tune the controller?

Thanks for you reply. I had the doubts when I was reviewing a design. I have no access to the controller or change the design at this moment. The hot fluid is hot water only.

The problem I mentioned was predicted with my own experience. As what I know, the control loop is suitable for fine tuning rather than large range fluctuation; it cannot take the operation parameter to the target immediately even though by tuning the parameters of the controller (say PID tuning) to increase the response speed. Even if this function is achieved at the moment of startup by tuning the controller, the control loop will over-react during normal operation. What I know, the controller parameters are set properly for normal condition. The controllers need be manually operated if the corresponding process parameters need to vary a lot or change fast during startup, shutdown or abnormal sitaution.

As what I know, for this kind of loop, the proper way is to manually operate the control loop. For example, the target temperature is 50C, the operator should manually slightly open the bypass control valve when introducing the hot fluid into the unit, and manually adjust the TV opening until the temperature is steady around 50C, then switch the controller to AUTO mode. Obviously, this operation will be annoying since the unit will start and stop frequently, so the designer wanted to keep AUTO control all the time. But I think it is not easy to achieve the design purpose, even if the TV could be in cooling water circuit.

To certain extent, this may be a typical issue for intermittent process loop control. I want to learn some experience for automatically maintaining the process parameter quickly at desired value during each startup in intermittent operation unit. I am wondering if there is proper process control strategy or instrument control technology available.
Thanks.

[JMW]

It seems to me this circuit is incomplete.
I would suggest that ZCV should have a status signal (e.g. limit switch) that is input into the controller such that when ZCV is closed the control valves position(s) are frozen at the last good value.

I would suggest a control valve in the cold water supply to the heat exchanger and a temperature sensor in the hot fluid stream immediately on the outlet to the heat exchanger; a second controller uses this temperature measurement to modulate the cold water flow to the heat exchanger. This is fairly slow responding control compared to the existing temperature sensor and control valve which should be capable of fast response and its purpose is to compensate for any sustained and significant change in the hot fluid temperature at the inlet t the system.
The set point temperature would be below the target temperature at TC and sufficiently below such that both TV and CV (cold water control valve) are ideally mid range and in their optimum control positions under normal conditions.

The problem of maintaining stable temperatures is less to do with the control valve response than the ability of the heat exchanger to respond rapidly to a fast temperature transient.

So, start with the system when the fluid in the system is stationary and cold i.e. with ZCV closed.

In this case the existing temperature controller would tend to have valve TV full open. But ideally a limit swicth on ZCV would have enabled the valve position to have been frozen.

Now, if when ZCV is opened the control outputs are maintained at their last good control condition until either a set time elapses or until the temperature comes within certain limits whereupon full auto control is restored. i.e. this can be simply automated.

Now the hot flow gets a chance to purge the cold fluid from the system and for the heat exchanger to optimally cool the portion of the flow through it before TV is put under automatic control and thus when auto control is restored the various valves are already at or near the optimum control condition and the hot flow arriving at TC should be near optimum anyway.

The purpose of a second temperature sensor and the control loop regulating colw water flow is to enable the heat exchanger to respond to trends that would otherwise take TV away from its optimum control position which could happen if the hot fluid temperature changes significantly for any reason which would cause a significant control response.

Now, when ZCV closes the limit switch should act to freeze the control valves in the optimum control position. How this is implemented depends on the valve actuators. Are they air open and air to close? or spring close? Is it simply a matter of interrupting the air supply or does it require that the temperature controllers have the facility to be operated as required in response to the limit switch input?

[Zauberberg]

Montreal,

Refer to excellent article written by Walter Driedger. I believe you'll find there everything you need: CONTROLLING SHELL AND TUBE EXCHANGERS

Throttling CW is generally not a good idea; exceeding 45C cooling water outlet temperature will cause severe and accelerated fouling problems.

Good luck,

[JMW]

Nice article, thanks for the link.

Assume the hot stream is 150degC, everything is under control and the ZCV valve is at its optimum control position e.g. mid way.

Assume that the heat exchanger sizing is such that the outlet temperature of the cooling medium is within limits and that a suitable flow split has been achieved. It may be that the temperature of the process fluid emerging from the heat exchanger is 40degC and that the proportion of the flow cooled to 40degC is such that the combined flow is at 50deg C.

Now assume that the process fluid temperature increases by 5deg.
The control valve modulates to bypass more fluid through the heat exchanger.
For a constant flow of water coolant the temperature of the emergent process fluid will now be not 40degC but slightly higher. Hence the flow proportion must be increased more to accommodate the higher temperature of the emergent stream than if the temperature here were to remain constant.

If the temperature variation in the process stream is a transient, then this doesn't matter too much because the situation will recover and it is exactly this type change that the bypass method is well suited to deal with i.e. a fast and effective control response to the temperature change which allows the output temperature of the process stream to be maintained with significant precision even if there are significant transients in the process temperature.

However, if this is not a transient, what we see is a more significant control valve change than would be desirable and we can also suspect that a further change in the process stream temperature may quickly take the ZCV control valve out of its optimum control band while at the same time the rising temperature of the process fluid from the heat exchanger exacerbates the problem.

By including a control valve of the control of the cooling medium and setting a suitable set point e.g. 40degC. then as a temperature change arises the immediate response of the system is to manage the flow split by modulating zcv. This is a fast response that helps provide good stability in the process discharge temperature.

The second temperature sensor, monitoring the process stream temperature as it emerges from the heat exchanger but before it recombines with the bypass flow allows a comparatively slow response. This will not be much affected by transients but for a sustained trend in temperature, it allows the coolant to be regulated which has the effect of increasing the cooling which reduces the flow rate through the heat exchanger which in turn brings ZCV back toward its optimum control setting.

If there were no bypass then the only approach to temperature control would be either to throttle the process flow or throttle the coolant. In either case precise temperature control is not possible due to the thermal lag in the heat exchanger. Note that in the article the preference is given to modulating the heat exchange medium

[JLMONTREAL]

Dear all

Thanks for all you input. As there is no chance to make many modification, finally we decided to link the opening of the TV with the ZCV (on/off valve). So, when ZCV is off, TV will close, and if it is on, the TV will open at certain %( need adjustment during commisioning) to avoid large fluctuation to the temperature and the control valve.

This function can be implemented by PLC programming. Althogh it is not very professional, but for this specific project, it is acceptable.

Thanks again.