Temperature problems with large commercial hot water system

[cr0ft]

Hi all. As most of you will know I've gone back on the cards now and am now fault-finding one of our commercial hot water systems in a 5,800sq.m building. It's a large building on a hospital site.

The hot water system is an old system! Heating is run from a steam feed from the main hospital and that heats the water in 2 very large calorifiers (3000 litres each). The water is heated to 65C in these and is then pulled around the system with the standard 2x pumps (one on, one standby) on the hot water return.

The hot water flow pipework is in 54mm and quickly splits off into 6x 35mm 'risers' feeding different parts of the building (6 branches really, all pipes run through between the 2 floors). The issue is that hot water return temperatures furthest from the boiler on some risers are down to 37-42C, and there is a legionella problem in the system. No shock there! When you open hot taps in the affected areas the temperature will then rise by up to 5C before dropping back as soon as the tap is closed. In the affected areas the water temperature doesn't even get up to 50C when running the hot tap.

From having inspected two risers (they are above suspended ceilings) I see no balancing/gate valves on the main return pipework. What I do find odd though is that every individual tap has it's own tee'd off hot water return pipework. The end result is there are loads of loops for the water to follow to get back to the calorifiers on each riser.

I suspect that the hot water is following the shortest routes back to the calorifiers and there are large stagnating areas where recirculation is not occurring. I don't see the point in the hot water return pipework to individual taps (they are all within 5m or so from the main pipework) and feel it complicates the system. My plan is to remove all of them and fit automatic balancing valves on the return leg of each riser in order to properly balance the system - something like the Honeywell Kombi 4.

All the schematics I see online show taps etc connected to the hot water flow pipework only, I'm assuming connecting them all to flow and return is wrong and is going to cause this issue I have surely?

Anyone else got any thoughts? Have spoken to our commercial plumber we contract in and he agrees with me but wanted to get other people's ideas before briefing management on Friday!

Thanks a lot in advance for all advice.

 

[ShaunCorbs]

Yup will recommend fitting isolation valves to all risers in one shift first then and may as well fit a balancing valve to returns on each at the same time.

Then I can work on a section at a time which will make it a lot more realistic.

And drain valves

But on the src and hot risers you will need to remove all the tees for the return circuit bar the furthest one

 

[Dkdc71]

Are the pumps sufficient from an original system with extension's and extra wings added to hospitals over the years might be undersized just a suggestion

 

[cr0ft]

Yes I'm going to recommend fitting some variable speed pumps in place of the existing ones too. Without balancing though won't they just push water round the easiest loops faster?

 

[Vee]

Somehow I think tackling the pumps first would be wise......

 

[cr0ft]

Been back to monitor system today and all return legs are getting up to 51-53 at the furthest points from the plant room.

On further investigation the pumps haven't been swapped for ages as the maintenance team haven't been shown how to do it.

Swapped the pumps, re-measured some temperatures and they are up by 4C.

I think we will be replacing both pumps first and seeing how we get on from there.

Out of curiosity, is it possible for a pump to intermittently pump at a lower speed?? I only ask as this system was sitting at 42C on furthest return pipework yesterday and the day before. Incredibly strange!!

 

[ShaunCorbs]

No as the only adjustment if any would be a selector on the side

And sounds like it's on it's way out it's not bms connected is it ?

 

[cr0ft]

Hi, it is but it's an incredibly old BMS system. What are you thinking? Will I be able to get pressure drop through it or something useful?

Just to be clear, here's an attachment which illustrates how the whole system is pipe (I've obviously only drawn a little bit of it). Had a chance to go through the whole building today tracing the main pipework. Turns out there are only 3 risers (duff information!).

Given that switching from one pump to the other gives me a 4C difference in temperature drop from calorifier to end of return legs (57 vs 53 today at end of return legs) we need one new pump I reckon to replace the more worn one.

I'm assuming the thermostatic balancing valves can stick like TBVs over time and so they are starting to play up. It's the only way I can explain temperatures of 42C and less at the end of the return legs earlier this week. Am going to test them more carefully tomorrow morning.

What I don't understand is how just placing TBVs at the end of each riser makes much difference though. Surely the water is just going to be running through all of the individual tap branch flows and returns before it gets to the end.. I would have thought each tap return would need its own TBV as well in order for the system to run efficiently. I'd also have thought we would need variable speed pumps so they match this demand but as far as I can see they just run at one speed.

I'm going with suggesting one new pump and fitting new TBVs atm unless anything else comes up tomorrow.

 

[ShaunCorbs]

I would if you could run the src on hand and not auto for a day

And best to replace the main and back up as you don't want mis match pump sets

The drawing to me looks like someone has teed into the src on the first 3 sinks (or x amount) after it was installed as the balancing valve is doing more harm that good where it is,

Could understand if there were balancing valves on the first 3 sinks (or x amount) as this would stop short cycling

As for pump spec wouldn't go ver speed more hassle than there worth esp in a hospital fixed speed less to go wrong bullet proof

Best speak to grundfos or dab in relation to sizing if specs arnt on the pump

 

[cr0ft]

Hi again and thanks for all of the info. Basically nearly all sinks have their own dedicated return pipes. There are gate valves on both flows and returns to all sinks. My assumption is that originally these would have been used to balance the returns properly. Over time the contractors in the building have used them as isolation valves then just opened them fully...

Currently there is a balancing valve at the end of each of the three return lines. There are no other automatic balancing valves on the system.

I'd imagine grundfos will visit and specify us two pumps that will retrofit the existing ones?

Our current setup is 2 separate pumps that are manually switched over each week. It's not an automatic process in this building hence my thoughts were change one pump first (they are bloody expensive as I'm sure you know), run it for a week or two and see the effect on temperatures. This is (I believe) the only building in our estate that has to have the pumps manually switched weekly. I also know for certain that our maintenance staff haven't been doing it so one has been running and the other not being used for ages. This is also the only building we have an issue with the hot water system in. Perhaps there's a clue there!

If there were no flow and returns to individual sinks etc, the balancing valve would work a treat at the end of each main return branch. As it is, it will be shutting the return off and encouraging short cycling round the earlier taps right?

I found out today there was a load of work done on the plumbing to the wards a few years back and we've had these problems with temperature ever since. I suspect the gate valves were used as isolation valves and now the whole system is out of balance. On some days it works great and on others it doesn't...

Very hard to prove as today is was working a lot better but I can't see what else it could be.

 

[ShaunCorbs]

Yep and they get messed with there's your problem how come there not on your drawings?

They only want to be just cracked open just enough for a trickle ( about just less than a 1/4 open)

Yep without working it all out your looking about 1lps (ish) flow rate max from each sink / pump manufacturer might work this out for you if you ask nicely hell if you can find the plans from when it was installed it will be listed on them

Yes and no you want to have a site meeting with a rep

And yes should be other than a flange change

Yea spread the wear

Yes you would have hot hot water on your main lines/ runs it would be the tees off to each sink that would not be recirculating so would chill

 

[Sparkgap]

Recent hospital jobs we've done they've insisted on virtually no deadlegs, taking the return from right next to the basin/sink/whatever isolating valve. How far are the basins off the hot main and are they regularly used? If not far there could be a case for doing away with the individual returns. Depends on the authority.
I'm assuming the pipe insulation is okay and you're not losing too much heat from the pipes. I'd look at (1) pump rating, make sure you've plenty of flow/head in reserve, (2) sorting out the l/s valves on each return from the basins and perhaps put in a ballofix or similar next to them so anyone doing work on the circuit isn't hopefully tempted to use the l/s for isolation.