3 storey hostel, shower temperature fluctuations

[KesTheHammer]

Hi

I inherited this project from an engineer who left the firm, so I'm picking it up cold.

A quick overview of the system:

268 bed hostel, 3 storeys. 28000 litre cold water storage tank on third floor, heat pumps on a roof slab, 2x 6000 litre hot water storage tanks in the same plantroom as the cold water tank. The whole system is pressurized from the booster pump drawing from the tank. Pump is on a VSD, duty of ~ 450 l/min @ 3.5 bar, but is controlling the pressure at 4 bar after some complaints about water pressure - the pump cycles quite drastically - shuts on and off again about every 10-20 seconds.

The hot water reticulation is OK - the system is not balanced wonderfully, but there is enough draw-off that it seems to work OK. - I would have preferred a more balanced system, but the contract is over so we can't install thermostatic valves now.

The system seemed to be running decently - although apparently the plumber received a couple of call outs that we were not informed of. Then a power failure happened and the hot water circulation pumps burnt out when the power came back on.

The system wasn't designed with a power failure in mind, and the course of action was to manually open the bypass valve, this caused the problem that there was no water, and students still opened the taps (wanting water) and letting air into the system. Once the power came on, the circulation pumps were running dry and burnt out.

About a month ago the manual bypass valve was replaced with a non-return valve and the pressure reducing valves were set 3 bar. The idea was that the municipal main water pressure would take over in case of a power failure.

Since that "fix" was implemented, there are complaints about the water temperatures in the showers fluctuating suddenly and scalding the students.

I need to figure out why?

I am inclined to suspect the pump because of the cycling and I don't think it is being controlled well by the VSD, but surely that will provide fluctuations in pressure, but the balance should remain constant.

Another theory that I have is that there might be a blockage in the cold water lines, which effectively reduces the diameter of the line, and when there is additional cold water drawoff, the cold water line is affected too heavily and the shower becomes unbalanced.

Any other theories?



Kestell Laurie
South Africa

 

[PEDARRIN2]

Are pressure balanced shower valves installed? It sounds like they were not, if you are having some pressure problems and getting temperature fluctuations.

You mentioned no thermostatic valves. How are you controlling the output temperature from the storage tanks? If you have the ability to control the temperature, you could set it lower.

Unless the shower valve is a two handle type, there is typically a limit stop on them so they can only reach a certain temperature. If present, what temperature was it set to?

 

[LittleInch]

You have given a decent description, but is still very difficult to follow without a schematic sketch of your arrangement, with pressures and flows marked on it etc. I can't work out where your bypass valve is and what this new NRV and PRV are doing and where the municipal supply is or why your big cold water tank didn't continue to feed during your power outage. Remember we only have the info you give us so can't see what you can see in terms of diagrams and knowledge of the system.

It does sound like your VFD is either oversized or not operating properly if it's busy cycling as much as that, but without some sort of diagram we're all just guessing.

My motto: Learn something new every day

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

 

[KesTheHammer]

I went to site today and managed some more information.

First, PEDARRIN2:

The showers are controlled by 2 taps - not a mixing valve. The water in the tank was about 58 °C, but it never drops below 55°C, and rises to 60°C - so the hot water is fairly constant temperature and is controlled by the heat pump.

At the bottom is a basic detail of how the showers are piped.

I witnessed the problem first hand. When the hot water is running alone, there is virtually no temperature fluctuation, and I estimate the temperature to have been about 40-42°C (I could hold my hand there, but it would probably have been too hot for my body). The water pressure feels fairly constant. Once I open the cold water tap to try and control the water temperature a bit lower the problem starts to come up.

Firstly - the cold water tap is only required to be slightly open (don't think it's 10%) to achieve the desired temperature, then when there is additional cold water draw-off (opened the basin), the temperature rises quickly and then suddenly falls down all the way to probably 20°C. This all happens in less than 30 seconds. The temperature keeps oscillating without any change in the valves - I can imagine that if you are showering, you will try to compensate to the cold temperature by closing the cold tap.

Now, moving up to the plant room - I did nothing to control the draw-off, it was just the regular draw-off from the students using the showers etc.

The pressure at the pressure gauge after the pump oscillates like crazy - the pressure would be about 410 kPa and gradually drop to about 390, then quite a sudden drop to about 325 kPa (assuming draw-off), the pump starts to compensate and within seconds it jumps to about 450 kPa.

The pump makes a terrible rattling noise at low speed. The plumber said that the bearing vibrates at low frequencies.

I have suggested that we change over back to the municipal main supply (i.e switch off the pump and increase the PRV setpoint to 4 bar), and see if we have any problems over the weekend. If the problem is solved, we know that the problem is at the pump. If, however, it is not, then I'm stumped.


Kestell Laurie
South Africa

 

[LittleInch]

It sounds to me like your VFD control loop is not sufficiently damped and also a single unit trying to cope with probably a 90% turndown ration (max flow to min flow) is just far too much if indeed it is a single unit. If the pressure is dropping and the pump isn't responding in time, it looks like the cold water feed is upsetting the control loop as the pressure is now being affected by a secondary input instead of just the booster pump.

I would try the following:

Reduce the pressure set point for the incoming cold supply to reduce the impact of this on the system, say 2bar or the minimum you need. If the lights go out there won't be many showers going...
Find a controls engineer and start to tweak the PID control loop to give a slower response and slower ramp up and ramp down of the pump speed
Consider adding a bladder system to even out the pressure swings
Run the system with the bypass closed, but monitored so that in the event of power outage you still have water or put a Fail Open actuator on the bypass valve

If your municipal supply is capable of delivering the required pressure what do you need the tank and pump for??

My motto: Learn something new every day

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

 

[KesTheHammer]

I agree that the pump should have been two smaller pumps instead.

There is a small bladder system in, but it makes almost no difference, since it's only about 2 seconds of flow worth.

The reason for having the pump in is two-fold - first, there is a requirement for fire purposes as well as back-up purposes.

I suggested a fail-open actuator on the bypass valve - but it was argued that the non-return would be sufficient since the Pressure Reducers will keep it from flowing when it is not required. I should not have yielded on that one.

Kestell Laurie
South Africa

 

[LittleInch]

The other question is whether you actually need 450l/min or you could survive with a bit less and hence skim or replace the impellor with a smaller one so that the ramp up and down of the VFD isn't quite as dramatic.

Your bladder tank is very small so not much benefit there, but long term this should be a lot bigger to smooth out your pressure fluctuations

Have you tried setting the VSD to a fixed speed or controlling in manual?

what does the controls guy say?

My motto: Learn something new every day

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

 

[PEDARRIN2]

Once you get the pump in order, I would also recommend lowering the hot water temperature. At 60 C, you can get permanent damage from burns in not a lot of time.

With a two handle shower valve, scalding is a definite possibility, especially if it doesn't take much cold water to reach shower temperature.

 

[KesTheHammer]

Like I said previously, I inherited the project.

a quick order of magnitude calculation, using my regular spreadsheet shows that 300 l/min would probably have sufficed.

But in any event. We will be testing the system on municipal only first, then on pump only and if it works well on both, we will probably install a fail open valve.

What I find interesting though, and what my colleague keeps getting back to is, that surely if there is a pressure fluctuation, the hot and cold water should remain balanced because it is served from the same source (whether it be municipal or pump).

Kestell Laurie
South Africa

 

[LittleInch]

I was wondering about that, but I think the issue is that for the cold supply, the flow is much less in the shower and hence more of the pressure drop is taken across the tap valve compared to the hot supply where more of the pressure drop is taken in the piping and becomes less sensitive to pressure fluctuations at the start point.

Those big hot water storage tanks could also act like additional bladders and maybe smooth out the pressure fluctuations more than the cold supply.

If you allowed the pressure to vary slowly then there may be less of an effect, but with a surging system all sorts of other effects come into play.

My motto: Learn something new every day

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

 

[PEDARRIN2]

Domestic water systems become unbalanced from pressure fluctuations all the time, whether served from a pump or from municipal source. Even on balanced systems, if you flush a toilet, the cold water to showers will decrease enough to cause a significant, but momentary temperature increase.

You may solve the pressure issue, but with two handle shower valves, you will always have a scald concern with the elevated hot water temperature.

 

[KesTheHammer]

Ok, after the weekend, the pumps were off, and the system is no better. I think the pipe sizes to the basins and the WC's are too large and aids the problem. Also the basins and the WC's are upstream from the showers.

Another theory that kind of ties in with the tanks acting like bladders, is all the pipes are monolayered plastic, and the hot water pipes act more like bladders than what the cold water pipes do. I've theorized that in my head, but I just can't get myself to believe it.

The table below gives the corresponding copper diameter:
Copper Plastic
15 mm 20 mm
22 mm 25 mm
28 mm 32 mm
35 mm 40 mm
42 mm 50 mm
54 mm 63 mm


Kestell Laurie
South Africa

 

[KesTheHammer]

I have spent the last couple of weeks on this problem and it seems that the effect is the classical draw-off effect.

Where a shower is running and a WC/basin on the same line draws off water, the shower gets less water and the balance goes off.

What confused me at first was that the pipes seem to be sized sufficiently to handle the flow.

The problem is that the water saving showerhead made the fluctuation worse.

The calculations below are only of an illustrative nature, but it shows that when your showerhead has a big dP, a pressure fluctuation on your supply line causes a much greater fluctuation than when it has a lower dP.

We adjusted the angle valves at the WC's and basins and that greatly reduced the problems, and the 25/27 votes on a facebook poll agreed that the problem is solved.

Kestell Laurie
South Africa