Pumps 1

[Isaacnewton4]

I trust you are well.

I have just started at this big boiler making company and there is a problem we are trying to fix with regards to water level in the boiler. We are looking into installing a differential pressure cell across the pump, such that the head of the pump can be changed as the boiler pressure modulates. This is to ensure that pump head is always greater than the boiler pressure so that there is flow at all times when required. I am working on a 11bar boiler as the test case. Feedwater tank is usually 3m tall. I do not know how to proceed from now on. I will appreciate any help on this

Regards

IsaacNewton4

 

[ashtree]

In the very simplest terms you are going to need a pump that can operate over a wide range of pressures but needs to be able to deliver the flow you need up to and above the 11 bar boiler pressure. That is not an excessively high head for pumping so that should be able to be accommodated with out a problem.

But what you also need to know is how much water flow you require and at what pressure, particularly if you are building this rig to service different size boilers. In order to select a pump at a minimum you must have a flow at a particular head or pressure specified.

I don't think you need to have a differential cell but merely a pressure and preferably a flow sensor such that the pressure and flow and the speed of the pump varied as required to suit the change in operating pressure.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[LittleInch]

First fully establish your requirements ( flow range and pressure range). What might seem a big swing to you may not to a piping and pump designer.

Also think about how you control this flow. Normally boiler feed pumps work on maintaining a fixed level in the boiler. Do you intend to maintain a fixed level, hence a possible wide swing in flow or start at a low level then inject water to a higher level and switch off? Makes a big difference.

For a single speed centrifugal pump you're probably looking at a range in flow of around 40 - 110% of rated duty at a pressure which could vary by +/-10% from your rated duty ( i.e. "normal" duty)

If your flows or pressures are wider than this then you need to look at two pumps in parallel or possibly some sort of VFD, but these are normally not very good for boiler feed pumps.

Or look at some sort of Positive displacement pump ( piston, screw etc) where flowrange is more like 30-110% but pressure can be 0-150% of rated at vitually any flow.

Feed water inlet pressure is about 0.3 bar so is fairly negligible in the grand scheme of things.

you've chosen a very special name for a handle - I hope you're up to it... welcome to ET, just make sure you read my strapline.

LI

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[quark]

This is what all boiler feed pumps manage to do. When your steam pressure is reaching set point, there is less steam consumption and thus you require lesser feed water flowrate. So, higher flow rates require lesser head and vice versa. I am unable to understand why feed water pressure is less than steam pressure. Please elaborate your problem.

 

[Isaacnewton4]

Thanks guys for your input.

Our philosophy is that once the differential pressure across the pump is found we can then use the signal from the DP cell through a controller and a VFD to make the pump head to be always greater than the boiler operating pressure hence ensuring flow. We also have a flow monitor that monitors flow to make sure that flow is happening when it is supposed to happen.

LI we have boilers operating from 11bar-20bar. We would then want the DP cell to measure up to 30bar. We intent to mantain it at a fixed level. We also have two pumps running in parallel.

Kind Regards

IsaacNewton4

 

[Isaacnewton4]

Thanks quark. The problem is our whole water level regulation control algorithm is not working right. We cant keep the water fixed at the average water level however the water level is modulating about the required set point. We understand the superiority of 3 element control but then we trying to do 2 element control since it is cheaper.

Kind Regards

IsaacNewton4

 

[LittleInch]

Ok, like I say, more details, better responses.

11 to 20 bar is a massive pressure range for parallel centrifugal pumps if you're not controlling them via a control valve.

How exactly are you thinking or doing the control "such that the head of the pump can be changed as the boiler pressure modulates." VFD? Control valve? different pumps?

If I was you I would just get pumps suitable for max pressure then install a control valve downstream to regulate flow and pressure accordingly. Simple, cheap, but you loose a bit of energy when operating at 11 bar.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Isaacnewton4]

We will use a VFD and a PI Controller for varying the head of the pump. We do not want to change our pumps because of cost constraints unfortunately.

 

[LittleInch]

I missed your earlier post about it being a VFD control, sorry.

So you appear to have a system which works, but has some level of liquid level variation. how much level variation? does it cause issues with the boiler?

When you say in your post above "We intent to mantain it at a fixed level" - Intend to maintain what? the differential pressure or the liquid level?
Why 2 pumps in parallel?

also what is a "flow monitor?" A flow meter or a simple flow / no flow device?

You say you don't know how to proceed, but what don't you know? You seem to have a working system so what is the issue here?

This is why it's best to describe your system in the original post, preferably with a diagram or two ( can be sketches) and clearly state what the parameters are, what you've tried and what your issue is. Drip feeding information is frustrating for those of us who can't see your design...

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[quark]

Yes, more description is needed. If you operate the boiler at 30 bar cut off and 16 bar cut in, and if you do speed control, you will have about 28% flowrate at 16 bar. When you are approaching cut off pressure, your consumption comes down. When you are at cut in, your consumption is maximum and reducing the flow at this point is counterintuitive. I would prefer LI’s suggestion. Just out of curiosity, why such wide pressure band.

 

[LittleInch]

VFDs can make a lot of sense in a number of systems, but boiler feed pumps aren't usually one of them. Especially if the VFD is being used solely to control flow the problem is that the static pressure / head element is often > 98% of the discharge head required with very little frictional losses. Hence if your boiler was at say 15.0 barg, at 14.9 barg discharge pressure you have no flow, 15.0 a dribble, 15.1 some flow and 15.2 lots of flow. A VFD for a pump with an 11 to 20 bar range is nowhere near accurate enough not to continually oscillate between too much flow and not enough.

Fix the VFD to a static high rpm suitable for 20 bar and then add a control valve controlling on level. Once you tune the PID controller it will be rock solid on level.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

Could your level control scheme at the boiler not be working properly because of steam condensation in the level dp cell's vapor leg - how have you set this up at the moment? This vapor leg / impulse line leading to the dp cell should be remote sealed to avoid errors in level readout.

 

[Isaacnewton4]

Thanks guys. I will try and revert to all your replies.

The first error I made in explaining is that we have 11 bar boilers, 14 bar, 17 bar and 20 bar boilers, not that the pressure in the boiler is designed to vary from 11-20 bar. So we intent to keep the liquid level of water in the boiler constant. We use two pumps but only one will be running at a time the other is for standby. Then the flow monitor is a flow switch which detects whenever there is flow. The main issue we have now is that the water is oscillating about the required setpoint whereas my boss wants the water to remain solid level at the required setpoint.

Thanks

IsaacNewton4

 

[LittleInch]

That makes a bit more sense, but the same comment applies - VFD control for boiler feed pump is a blunt tool because the VFD and pump simply can't adjust to the degree you need. I don't know what the granularity is for a VFD, but if its 1 Htz that will make a difference of fractions of a bar which will vary your flow a lot, hence the hunting on water level, because the static pressure is such a high percent of the pump output so flow varies very a lot and fast for only a very small increase in DP.

Just set the DP cell to say 3 bar above boiler pressure and add a control valve ( 3 bar should be enough) and then control on level using the control valve. Once you sort out the PID loop control it will get you much much better control over your level control.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[quark]

So, you have different boilers at different set pressures with common feed pumping system. What are the capacities of your boilers?

 

[Guac]

Are there issues with running the pumps by a float switch into the boiler? The pumps would turn on/off at desired set points and then run them off an off-delay relay to prevent a bouncing/oscillating float from making the pumps stutter on/off. This is how I have seen boiler feed water pumps.

-Blake

"If it rotates, it vibrates."