how to prevent temperature increase in pump system 1

[diapensiya]

Hello,

I have been working on a system which consists of a tank (at atmospheric pressure), centrifugal pump (normal operation flow rate is within 28-46 m3/h, during startup up to 100 m3/h) , and a column (around 2.5 barg)as end destination. Before the column, there is an heat exchanger which has maximum design pressure of 8 barg. Over the pump, there is a spillback to the suction of the pump. In this spillback there is a pressure control valve, which is usually set to 7 barg and to 3 barg during startup to prevent pressure spikes, damage to the heat exchanger. Therefore, this control is essential. In order to prevent overheating due to the spillback through the pump, I have changed the recirculation line to be through the tank with a 4 in line. However, now the problem is that the resistance through the pressure control valve creates pressure drop and thus the liquid cannot go up to the tank (elevated over 8m from the pump). It changes state.

Can you suggest me other solutions to prevent temperature increase through this pump system ?

 

[LittleInch]

Can you sketch out your system as there's something not right here. If you have 3 bar available at your pump discharge then you should be able to flow to a tank 8m above you. What is the fluid and what is the fluid level above the pump, that's what you need to know. However for a centrifugal pump this sounds like a big pressure variation (7 to 3 barg). Is the control valve controlling on pressure or is it just being manually set to percent open?

Also include pump curve if you have it. It sounds a bit to me like you're trying to fix something the wrong way. Why are you not controlling the pump discharge pressure on start-up. you only really need a "spill back" line if you're operating at flows less than about 25 to 30% of the rated duty. Are you?

More details will get you a better answer.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[diapensiya]

Thank you for the reply.I am certain that the system is right. Maybe the information I gave is not sufficient.
There is 3 barg available at the discharge of the pump. This pressure is set by the pressure control valve (It controls on pressure) at the recycle loop. It means inlet to this valve is 3 barg and outlet is around 0.8 barg depending on the flow. 0.8 barg is not enough for pumping a tank 8 m above. 3 barg stream goes to the heat exchanger and to the end users.
I need at least 0.93 bar to pump it to the tank. The specific gravity is 1.2 and vapor pressure is 0.02 bar.
I need a spillback to prevent the pump running dry when there is no flow to the users and to make sure that at the startup the pressure through the heat exchanger does not damage the equipment.

 

[LittleInch]

Sorry, but something here still doesn't make sense. Does the outlet from the control valve go only to the tank? If not then this is an issue because otherwise I can't see why there is no flow to your tank. you say the outlet of the valve is 0.8 barg depending on flow - where is this flow going to generate a DP across the valve?. If your pump pressure is being controlled to 3 bar by the re-circ line this is not normally a good idea.

Need to see a schematic (can be hand drawn) to work this out.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[diapensiya]

The red line is the line I want to use. The black is the current.

 

[HPost]

I see what you are trying to do, but it prompts more questions.

Why have you got a pump that can generate so much more head than you require?

It is not usual to have a centrifugal pump with a pressure regulator. You should use a pump that has slightly more head than you need for a given flow rate through the system. Any additional pressure is better and more accurately achieved by increasing the pressure on the inlet with air.

You are wasting energy by pumping water over the relief valve. You are making this much difficult than it needs to be.

Why is the flow higher on start up, is the pump motor on a star / delta starter?

Also, why are you trying to reduce the pressure when the flow is higher? That is just asking for trouble.

HPost CEng MIMechE

 

[diapensiya]

This is an old running system not my design. How would you solve the problem of temperature increase?
How would you start the pump without damaging the heat exchanger ?

 

[LittleInch]

Your valves are in the wrong place controlling on the wrong things.

From the left, the first valve on the red line is doing nothing and needs to be removed. You need to connect the red line to the black line and then delete the black line going back to the pump inlet.

The second valve should be removed and replaced at the tee connection ot the discharge by an Automatic recirculation valve (ARV) - look them up. This will relieve the flow back to the tank when flow past the tee is below a certain value. Currently I suspect your pump is going to the end of its curve when the control valve on the recirculation line is set to 3 bar, if it's normal operating point is higher than this. you might need to install an orifice plate on the line back to the tank to generate some back pressure to stop your pump going beyond it's normal operating range.

The third control valve should also have a pressure sensor downstream of it which can then control the pressure to your HX to whatever you want and control the flow to whatever you want. This is the key to better control of your system. The current configuration is just plain wrong.

If the flow through the pump is less than your pump minimum, flow will go back via the ARV and the recirculation line to the tank giving you a much bigger volume to heat up.

If you really really want to keep going as you are, then remove the firs valve on the left and just connect the control valve direct to the tank and remove the connection to the pump inlet. Currently this is where the flow will go as it is at a lower pressure than your tank so will heat up very quickly. However I would not recommend this as I think the flow through the pump on start up is much too high and is also causing excess heat to be generated.



My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[diapensiya]

I agree that first valve on the left should be removed. Currently it is kept closed, thats why it is in the sketch. And yes I also consider putting an orifice after the pump not to go beyond the operating range.

What I understand is that you suggest that ARV taking care of preventing dry running of the pump and the pressure and flow are controlled by the third valve before the heat exchanger. Is this correct ?
If so, could you please help me understand more; at the startup ARV will close the flow to the column and automatically open the bypass line to the tank. Then, when there is some flow to the column, ARV opens both lines till it reaches minimum flow. When the flow to the column excees the minimum flow for the pump, then ARV closes the bypass line and only opens the line to the column. Here the flow is set by the operator using the third valve on the right, which is a flow meter. The flow is determined by the process needs. So how will the pressure sensor function here ?

 

[LittleInch]

Basically you've got it. The ARV's close gradually to maintain the minimum flow so once some flow goes to the column, the remainder of the minimum flow goes back to the tank. The orifice I mentioned is downstream of the ARV on the line back to the tank to give the pump some back pressure when flowing at the minimum flow for the pump, not on the discharge line.

The third valve can normally undertake two or more functions, especially if it is a PLC controlled valve. The valve controller just takes the lowest of the inputs either from the pressure controller (initially set at 3 bar for start up then increasing to some other higher value) or the flow controller and controls accordingly. Its called a low selector block in control engineer speak. see simple sketch attached.

The problem with your current set up as I see it is that the previous designer was trying to control start up pressure by flowing large volumes back to the inlet and essentially running the pump beyond or at the end of its curve as this is the only way to reduce pressure in a centrifugal pump. Hence lots of flow, inefficient operation and more energy being dumped into the fluid. This is not the way to run centrifugal pumps, you control pressure directly in the output line by either a speed control on the pump or a control valve on the discharge line or by sizing your pump correctly in the first place... If you really need to you can introduce a re-circulation line for low / no flow conditions.

hope this helps.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[diapensiya]

Thank you for the sketch. Indeed it is where the start-up flow is.
To give a simple example of how PLC controlled valve acts:
Lets say
Min flow requirement = 11 m3/h
Set point of flow to column = 5 m3/h
So By-pass line flow = 6 m3/h
Total flow through the pump = 11 m3/h corresponds 7 barg in pump curve
If max pressure PLC valve should allow is 4 barg for Heat exchanger , then what does PLC valve do?
Let 7 barg with 5 m3/h or set 4 barg with say 85 m3/h ?

 

[LittleInch]

The valve will control on whatever is the limiting function. Your system can apparently cope with 5 m3/h to the column with only 3 bar available as per the current configuration. Question - At 5 m3/hr, what is the current pressure in the HX? If it is less than 3 bar then your PLC will control the valve on the 5m3/hr set point until you raise it higher. If it is more than 3 barg then you might need to raise your arbitrary 3 barg set point a bit higher.

I still don't understand why your 8 barg HX needs to be protected to 3 barg by this controller and there may be a bit of adjustment of some set points but if you find that by setting the d/s controller to 3 barg you are not getting 5 m3/hr, then set it a bit higher until you start to control on the flow.

Only you know your system and you've only given us a fraction of the details available, but this set up will allow you to optimise your current set-up, minimise heat build up and protect your pump from running at or beyond the end of its curve.

If you post the pump curve and some actual operating data it might be easier to comment further, but hope this explains a bit more.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[diapensiya]

I put some values on the curve in attachments.
Currently when the flow is set to 5 m3/h, it means that the discharge pressure is 3 barg by the discharge pressure control. The by-pass through the pump is huge, at the end of the curve. The heat exchanger sees 3 barg. The normal operation flow is around 25 m3/h. It is gradually increased from 5 to 25 m3/h. At the same time set point of the discharge pressure control is increased gradually from 3 to 7 barg. The reason is not break the tubes of the heat exchanger by sudden pressure increase.

What I did not understand is that how PLC controlled valve can satisfy both the flow and pressure demand. controlling one means controlling other isnt it?

 

[diapensiya]

Normal operation flow rate to the column is 25 m3/h. It makes the flow through the pump 45 m3/h. Rest is by pass.

 

[LittleInch]

Yes, flow and pressure are interlinked and as you note above, the current operation balances one against the other, i.e. if you left your pressure at 3 bar, you won't be able to increase flow beyond a certain amount so you increase the pressure until the valve controls on flow then increase the flow until you need to increase the pressure etc etc.

I thought that was how you were running this system which is just plain poor design and operation - I'm surprised your pump isn't complaining or tripping, but maybe it's just big and strong ;-)

The PLC automatically chooses the value which gives the lowest percent valve opening. As you are controlling on two inputs you gradually need to raise one or the other to allow the control to pass from one value to the other. It is a bit strange to get your head around, but it works in practice very well.

I still think your reason for the pressure control on start up is very odd, but that's your decision. what sort of surge / transient were you getting?

so long as the pump has more than it's minimum flow (around 30% of the "rated" flow") you don't need to recirculate any fluid - is this what you mean by "bypass" - the ARV should be closed by then.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[diapensiya]

The pump is tripping Sometimes the operation needs change. Then bypass is needed.
It is good to learn about different control valves. Thanks for the help and the information.

 

[ozmosis]

I'm no pump expert but always great to read an expert at work. Excellent advice Littleinch