Unbalanced thermostatic shower - fit PRV or PEV?

[James Carmichael]

Hi,

I've just fitted a Hansgrohe Raindance Select thermostatic mixer shower. With both hot and cold supplies running to it it gives water that is only "quite warm" (which is not warm enough for SWMBO to shower in it!) If I turn off the cold supply, it gives fully hot water.

Hot is supplied by an unvented cylinder (heated by a system boiler). This is the first item tee'd off from the incoming cold main. Feed pressure to the cylinder is reduced to 3.5bar.

Cold is supplied from the mains. When I last measured this it was around 4.5bar.

There is a lot of pipework between those pressure measurements and the shower. The hot also takes a bit more of a windy path through the house. My suspicion (unchecked) is that the pressure differential at the shower will be greater than the one bar measured at source.

I assume this pressure differential accounts for the lukewarm water issue. I'll check in the morning by turning a few other cold taps on, see if I can get the shower to right itself.

So, what to do? With practical local limitations in mind, options I can come up with are:

(1) Fit a PRV to the cold at source, immediately after the feed to the hot unvented cylinder, bring the cold feed to the whole house down to 3.5 bar, matched as best I can to the hot. Easy to do, but problem is it doesn't guarantee there still won't be a pressure differential at the shower given very different pipe runs.
(2) The PRC to the unvented cylinder actually has a balanced cold supply (unused). This would be a real faff to do, but have the whole house cold come off this. Same problem as (1), but I guess you can at least guarantee the thing is 100% balanced at source.
(3) Using that balanced feed just to feed the shower is not an option - would involve ripping up building fabric to plumb a new hot supply.
(4) Fit a PRV on the cold supply on the branch to the shower (so, only the shower cold feed is regulated down).
(5) Fit a pressure equalising valve on the hot and cold feeds to the shower.

I think I am favouring (5) - thoughts?

As always, many thanks for your help (and apologies for long post - trying to be thorough).

JC

 

[Doggit]

Have you got isolating valves on the shower cold flow. Turn that half down, and see if the situation improves. If it does, then decide whether it's going to be easier to fit a PRV, or a PEV, but it controlling the flow may allow the shower to work better.

 

[James Carmichael]

Hi Doggit,

Thanks for your reply. Yes, I have isolating lever valves on both hot and cold. I tried partially closing the cold before I posted but, as you can imagine, they're a bit of a blunt weapon - from 180 to 90 degrees, no difference, but from 90 degrees to 89.9 degrees is dramatic. Really, the best I could conclude from them is that by turning the cold right off the shower is hot. And, the hot flow is highest when the shower temperature is turned to hot, and negligible when it is set to cold. This leads me to believe that I have, at least, plumbed the thing the right way round!

I've turned on all the cold taps in the house (the new shower is the last item on the pipework) to reduce the cold pressure, and that makes a notable difference. So, I'm happy it's a pressure balancing thing. Having slept on it, I'm going to stick a PRV on the incoming mains, immediately after the feed to the hot tank. I can always dial it up past the actual delivered pressure if, for some reason, I later decide I want to switch it off.

Thanks

JC

 

[John D v2.0]

You need both sides off the prv to stop one dominating the other. If they're off different pressure it can take all the flow from one side until the residual pressure meets the set pressure of the other side.
Thermostatic would handle it better then a mixer but even they have limite

 

[James Carmichael]

Hi John,

Thanks. My plan is that both hot and cold will be regulated down to the same-ish pressures, albeit using two PRVs - the preset one already on the feed to the unvented hot cylinder and a new one I will add to the cold.

The pipework isn't really conducive to using the same PRV (I can't easily insert one on the cold before the tee off to feed the hot, IYSWIM).

I can't believe, though, that it can really be that sensitive? Yes, I could exactly match hot and cold pressures at some arbitrary point in my pipework - say, as it comes into the house. But, there's no guarantee that H+C will follow exactly the same course on the way to the appliance. My hot feed has many more bends in it and follows a longer route. On the other hand, my cold feed has more tees in it, to feed cold only things like outside taps and the washing machine. Bernouilli tells us all of these differences will affect pressure.

And, even more tellingly, people can turn on hot and cold taps at will, again affecting downstream pressure.

So, surely a thermostatic shower head must be designed to accommodate a reasonable difference in delivered H+C pressures without causing an unreasonable difference indelivered temperature?

Anyway, I've also treated myself to a pressure gauge, so just for fun I'll be seeing what the pressure differential at the shower actually is!

JC

 

[Doggit]

You're thinking along the right lines James. If the showers a good one, it should handle a slight difference in the pressures, so you may get away with one PRV, and if you can fit it somewhere in the pipework before it gets to the shower, then that'll keep a good cold pressure in the rest of the house. If you can't, then fitting it where it comes in to the house, will allow it to control both the hot and the cold water at the same time.

 

[Nozzle]

Frankly, 3.5bar to 4.5 bar isn't much different. You get a problem when you have cold that's mains fed and hot that is gravity fed, then the difference in pressure is much much more. I'll bet there's some temperature adjustment on your shower, there was on mine. It's set so the hotset it will go isn't "too hot" - I would put £1 on that the manufacturers ensure the set from the factory can't get too hot to burn a child or someone elderly on the first use. Read through the supplied notes and look at how you can change this range so the hottest it goes is a little warmer than you like it, so it normally sits at mid position.

Nozzle

 

[John D v2.0]

I can't believe, though, that it can really be that sensitive? Yes, I could exactly match hot and cold pressures at some arbitrary point in my pipework - say, as it comes into the house. But, there's no guarantee that H+C will follow exactly the same course on the way to the appliance. My hot feed has many more bends in it and follows a longer route. On the other hand, my cold feed has more tees in it, to feed cold only things like outside taps and the washing machine. Bernouilli tells us all of these differences will affect pressure.

It's more about the dynamic pressures upstream and downstream of a mix.
The static pressure is 4.5 or 3.5 however suppose the dynamic pressure after the mixer is 4bar due to good cold supply (sounds like yours certainly is) then the hot water won't move out of the pipe at all, so a freezing shower.
Turn down a valve in the cold and the static pressure will not change, however if you turn it almost off, your dynamic pressure will go below 3.5 at some point.
Now your system is in equilibrium, but now your cold pressure will easily vary above and below the hot pressure based on the cold flow rate demand.

Clearly in your case the thermostatic mixer should try to counteract the difference by closing the cold side of the mix almost all the way, but it's not really a comfortable situation for it to deal with.

Everything from one prv would guarantee an even mix, different length piping would have a proportional mix but still fine, but differently set pressures would cause an extremely unbalanced mix.

Also the lower the back pressure of your shower head the less problem there would be, but that shouldn't be a significant issue unless the shower is like needles or has a hous blockage.

 

[Madrab]

OP ... You need to use the balanced cold supply off of the combination group valve controlling the unvented, this is exactly why it is there. That ensures a balanced H&C feed to avoid this exact scenario. Anything else is just a bodge, is overcomplicating things and may or may not work properly long term.

 

[Hot&Cold]

Balanced cold connection all the time although purpose made flow restrictors can be fitted in many shower mixer valves.
Would have though the shower valve would cope temperature wise,flow and temperature need setting up during commissioning as per the supplied instructions.
Should be an easy fix

 

[James Carmichael]

Hi All,

@Nozzle & @Old&Cold - yes, there were commissioning instructions to alter the temperature range. I've followed those and 'turned everything up to 11'. There is no provision for fitting flow restrictors in the instructions. So, I'm fairly confident that there isn't a shower installation issue here.

@Madrab - Understood, but as per my original post that's not really a practical option. The shower in question here is in an ensuite. To get a new balanced cold main to it I'd have to rip up the tiled flooring in my main bathroom (or tear down the ceilings in the downstairs loo and scullery), lift the nicely fitted floor coverings in the hall, lift the nicely fitted floor coverings in three bedrooms and try to negotiate a safe route that allows me to notch the (roughly 12 metres horizontal run / 40 cm spacing ~ 30) joists I'd find along the way; joists that have been fairly liberally notched over the last 120 years already. Yup, can be done, but really not a job I fancy. In fact, I'd rather set my mind to inventing a time machine so I can go back ~5 years and install it in the past, when it would have been much easier to do

Anyway, measured the static pressures at the point of delivery to the shower. Cold = just over 4.5 bar. Hot = just under 3 bar. So, cold = 150% of hot. Additionally, the flows are different too. I wasn't quite accurate enough with my bucket and stopclock to get sensible litres/minute readings, but it's clear the flow on the cold far outstrips the hot.

All in all, a PRV will be going on the cold main tonight and I'll report back if it fixes it!

Again, thanks all for help so far.

 

[John D v2.0]

So, cold = 150% of hot.

True relative to atmospheric but not relevant unless you have atmospheric pressure at the output of the mixer., You need to measure relative to the dynamic (back) pressure of the shower head.

All in all, a PRV will be going on the cold main tonight and I'll report back if it fixes it!

Personally my money is on it fixing it, if it's close enough then it will even out through the pipe work. Do let us know!

 

[fezster]

Hi All,Understood, but as per my original post that's not really a practical option.

Hot is supplied by an unvented cylinder (heated by a system boiler). This is the first item tee'd off from the incoming cold main. Feed pressure to the cylinder is reduced to 3.5bar.

The PRV (with the balanced cold output) is on the input to the unvented. Simply move it to before the tee from the cold, so that both the hot and cold are after the PRV. This would be where it is normally sited in an installation.

All in all, a PRV will be going on the cold main tonight and I'll report back if it fixes it!

This should also do the trick (but an unnecessary additional PRV imo!).

 

[James Carmichael]

Hi All,

@fezster - good thinking on moving that balanced PRV, but I saw it too late ...

... because I've now fitted an additional PRV to the cold feed after the tee to the unvented tank. I guess that does mean I have a pipe, with a tee, with both sides of the tee having a PRV where, as fezster wrote, one PRV on the input side would do.

But, hey ho. It's worked. Toasty shower. And, what I hadn't realised until now as I'd always lived with it, is quite how much of a pain in the *rse 4.5bar of cold pressure is. All the other cold taps in the house are now behaving thermselves, no longer scatter gunning spray all over the place when I turn them full on.

Thanks all for your help, mystery solved, solution (maybe not ideal, maybe not optimised, but simple and operationally effective) identified, time for beer.

JC