pressure at the bottom of a tank 1

[Thermo01]

Hi guys,

I'm sure this is a really simple question for you guys but it's one that has me stumped.

We have a chilled water system with an open tank. The tank is about 4m high and is filled with water about 3m high. The primary pump that's connected to the chiller draws water out of the bottom of the tank, pushes through the chiller evaporator, and discharges into the top of the tank. The primary circuit is run in 150mm and the total piping is about 30m. 15m to the chiller and 15m back up to the top of the tank.

There are a couple of secondary circuits connected to the same open tank with separate pumps, that distribute this chilled water through the building.

We have replaced the primary pump on the system so had to drain the low level leg but after opening the valve at the bottom of the tank and filling the pipe at ground level, the pipe pressure (pump inlet) shows 0.1 bar. I would have expected this to show the head of the water in the tank, which is 3m or 0.3bar?

is it because it's an open system and the flow and return pipework balance out as someone has suggested?

Can anyone explain why this is?

 

[LittleInch]

I'm really not following this. - "so had to drain the low level leg but after opening the valve at the bottom of the tank and filling the pipe at ground level, "
What valve?
What "low level leg"
Filling the pipe ?? form where? with what?
Can you draw it up.

Is the 0.1 in static condition or running?

What span is the gauge?
Is it a proper gauge or a little 1" one?



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

 

[Thermo01]

Hi LittleInch,

I've attached a quick sketch of our setup, Hopefully, this makes things a bit clearer. I haven't included the secondary legs on the sketch because they aren't part of my question.

To be clear, This system has been running for several years. Our works have consisted of removing the existing pump and replacing with new.

Once the new pump was installed. The isolation valves were opened and the pressure reads 0.1bar on the inlet of the pump, yet the head in the tank is 3m.

Also the gauge is 0-4 bar - 4". I know this is quite a large scale and a 0-1bar would be better but this should still show 0.3bar.

 

[LittleInch]

Well on the basis that the pump is at the same level as the base of the tank, yes you should be getting 0.3 barg, but with a 0-4 bar guage I suspect you've either not zeroed the guage properly or the valve isn't open...

Most gauges have a hard stop at 0 so tru zero may be less than that. See if there is a zero adjustment screw, remove the guage then adjust it until it starts to move past zero then slowly back it off until you have a "true" zero barg

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

 

[Thermo01]

We've tried it with several gauges and they all ready the same.

I spoke with a colleague who explained that the reading should be 0 Bar. He used the analogy of a hose pipe bent in a 'U' shape. If it was filled up then the pressure would be 0 bar because the pressure on the left hand side of the hose and the pressure of the right hand side would balance each other out because they are open to atmosphere.

This doesn't make sense to me because the atmospheric pressure is the same both sides?

 

[TiCl4]

Thermo01,

If your colleague meant that pressure at the middle bottom of the "U" would be 0 barg, then he would be dead wrong. The gauge pressure would be the height of the liquid in the "U" relative to the pressure gauge. At any given height, the pressure differential from the right to left leg is 0, but the absolute pressure is nonzero. No matter the configuration of "legs", any liquid-filled volume will have a pressure at a given height that is equivalent to rho*g*h + headspace pressure.

If your colleague were right and the pressure at the middle of the bottom of the U were indeed 0 psig, then you should be able to drill a hole in the middle of the bottom of the "U" and not have any water flow out.

I would suspect a zeroing issue with the gauge. What is the height of the gauge relative to the height of the tank?

 

[LittleInch]

What's that weir plate?

Could you have 1m that side of the plate?

I think we need a better drawing / phot of the inside of the tank.

Do you have a good reading of the water level? Have you looked into the top of the tank?

where does the system top up?

You still haven't said what this "low level leg" is.

BTw your colleague is talking rubbish as eloquently explained by TiCL4.

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

 

[r13]

I think your colleague meant the gauge is at one end, on top of the u-tube - zero head differential.

 

[Thermo01]

TiCl4 - I can only assume that he was describing pressure differential although that isn't the question that I asked him. I will speak with him tomorrow.
It looks like the gauge may well be the problem but we have tried 3 brand new gauges and all of them only rise to 0.1 bar.

LittleInch - Definitely not. We have looked inside the tank and the water level has been measured at 3m on both sides of the plate.
The system tops up from a ball cock on the side automatically.
The 'low level leg' is the pipework as it leaves the tank at the bottom and goes into and out of the pump. The pressure gauges are mounted across the pump and read the same value.
I couldn't understand how the pressure reading wouldn't be 0.3 bar and whilst I didn't understand my colleague's explanation, I thought he may be correct because it was more of an explanation than I had, hence my post on here.

r13 - He must have done but he knows the hydraulic layout so I'm confused why he would talk about differential pressures when i was asking why was the pressure to the inlet of the pump (when off) so low.

Also as another question - Is it possible to know what the inlet pressure of the pump would be when running if the inlet pressure when off is 0.1 bar (assuming the gauge is reading correct. I imagine this would be to do with A/NPSH is this correct?

 

[r13]

I think your colleague was implying the 0.1 bar is the head differential between the discharge point (top of tank) and the bottom of the tank. Can it be?

 

[Thermo01]

How if the water is 3m in the tank?

 

[TiCl4]

Thermo01, calculating the NPSHa a priori requires a system analysis of the piping lengths, fittings, etc. to determine the flow and pressure drop associated with the recirculation loop. If you already know the normal duty point of the pump (the gpm when you are operating normally), then you would only need to evaluate the suction piping segment for pressure loss. This relates to the NSPHr on the pump curve, which is normally given in feet or meters of water absolute pressure.

Or you could simply get a absolute pressure gauge and directly measure the suction pressure.

What class pressure gauge is this? Normal pressure gauges are Class B gauges, with a 3-2-3% error, meaning the bottom and top quarter of span have 3%-of-span error, and the middle has 2%-of-span error. With a 0-4 bar gauge, that means a listed error of 0.12 bar. 0.1+0.12 = 0.22 bar, and you are expecting 0.3 bar.

It would only take a little bit of error in your measurement height differential and in the zero-point of the meter (zero-point error is NOT part of the span error!) to get the additional 0.07 bar error. Pressure gauges often come with a zero "box", and 0.07 bar difference in zero-point pressure would only be a 6 degree travel in the needle!

Do you have a pressure gauge with a smaller span?

 

[r13]

The head at discharge point is 0.4 bar, at the bottom of the tank is 0.3 bar....

 

[itsmoked]

Have you run the pump long enough to purge all air out of the loop?

Completely cut thru the gauge nonsense.

Turn off the gauge port.
Remove the gauge.
Screw in an adapter that goes from the gauge port size to whatever bigger 3 m of pipe you have laying about.

Then just look at the precise head that's present.

Alternative: Use clear tubing for a sight-glass affect.
Alternative: Use any flexible hose including garden hose and starting with it at the top of the tank lower it until you get the liquid level with the opening. (Standard liquid level style)



Keith Cress
kcress -

http://www.flaminsystems.com

 

[LittleInch]

Thermo,

You've got us all stumped here, so if you can either accurately measure everything in elevation (tank, nozzle, pump, and most importantly the actual guage height all relative to some fixed datum and/ or post a few photos then we might , just might get some insight, but it could easily be a gauge issue where it has some drag on the needle and mechanism at the low end.

You can calculate the frictional losses from the tank to the pump and then subtract that from the atmospheric pressure and the static height of the water over the elevation of the pump inlet. This will give you the inlet head/pressure at the pump.

Or buy a digital battery powered pressure transmitter/indicator...

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

 

[Thermo01]

Hi Guys,

Sorry for the late reply on this. It does look to be a gauge issue. I connected a digital gauge and I got a reading of 0.29 bar, which is how I understood it should be.

My colleague had tried several analog gauges and they all showed 0.1 bar.

Thanks for all of your help and suggestions on this, it really is appreciated.

 

[Compositepro]

Interesting that a few cheap gauges reading incorrectly can cause some engineers to question basic principles of physics.