Skip to content
View in the app

A better way to browse. Learn more.

BuildHub.org.uk

A full-screen app on your home screen with push notifications, badges and more.

To install this app on iOS and iPadOS
  1. Tap the Share icon in Safari
  2. Scroll the menu and tap Add to Home Screen.
  3. Tap Add in the top-right corner.
To install this app on Android
  1. Tap the 3-dot menu (⋮) in the top-right corner of the browser.
  2. Tap Add to Home screen or Install app.
  3. Confirm by tapping Install.

Domestic hot water tank located outside the house?

Featured Replies

Hi all,

 

Looking for some advice on the placement of a DHW cylinder and buffer as I’m in the middle of a ‘spirited discussion’ with the plumber about where the DHW cylinder should be placed.

 

The ASHP is going to be located behind the wall of the car port. Supply/return pipes will then enter the building as shown in the attached diagram, into the loft space above the guest bedroom (building is single storey).  The loft space will be within the heated envelope of the house.

 

The plumber is adamant that the DHW needs to be as close as possible to the ASHP to reduce heat loss.  As a result, he’s saying we should put it outside, next to the ASHP and run pipes into the building from there (location 1 on the diagram).

 

I understand the need to reduce heat loss but can’t help thinking that putting the DHW cylinder outside the envelope of the house is going to cause heat loss in itself, along with longer term maintenance issues due to damp/cold etc (it would be located within a wooden structure for basic protection but nothing like the protection it would get inside the house).

 

My thought is that it should go at location 2 – centrally within the house in the loft space.  This will help with even distribution of DHW around the house, is closer to the main water supply point and will be protected from the elements.

 

Can anyone see why we should mount the cylinder outside the house?  Any pros/cons I’ve missed?  Better solutions?

Screenshot 2026-11-16 at 14.45.43.png

You charge the hot water tank once maybe twice a day? The longer the pipework from the heatpump to the cylinder the more you pay each time you charge (have to warm the water in the pipes). But you also pay to warm the water in the pipework that feeds the taps/appliances and you use them far more than twice a day. So from my POV the tank should be located wherever it leads to the shortest runs to the taps (and is still convenient).

The insulation on the tanks is designed for use inside. I know putting them outside is done sometimes (at least in other, warmer, countries) but first time I've heard of it suggested here. If you were to do that, I would want to build an insulated enclosure for it, not just leave it in the elements.

Given what you've said I only see downsides to your plumbers proposal. Use high quality insulated pipework to feed the tank.

I'd side with you on this one tbh. A cylinder loses heat all day every day whether the heat pump is running or not, so if it's in the heated loft that standing loss goes into the house for most of the year. Stick it outside in a timber box and you're heating the garden, plus you've got frost protection to think about on a tank full of water with the valves and expansion vessel sat out in the damp.

The primary pipes from the ASHP only lose heat while it's actually charging the tank, once or twice a day like rick said, and if they're lagged properly a few extra metres up to the loft doesn't make much difference. As long as the heat pump's circulating pump can manage the run it's not abig deal, worth asking the heat pump maker what pipe size they want for that length.

Couple of practical things with the loft though. A full 250 litre cylinder plus a buffer is well over 300kg so it wants a proper platform spreading the load across several joists, ideally over a load bearing wall, so check that with whoever did your structure. If it's unvented you also need a route for the discharge pipe and tundish down to outside, and room to get at it all for the annual service.

Central is better for the hot taps anyway, short runs to the bathroom and kitchen save more over the years than the primary pipe loss

  • Author
4 minutes ago, JpManning said:

A full 250 litre cylinder plus a buffer is well over 300kg so it wants a proper platform spreading the load across several joists, ideally over a load bearing wall, so check that with whoever did your structure. If it's unvented you also need a route for the discharge pipe and tundish down to outside, and room to get at it all for the annual service.

If put in my preferred location the floor area below the tank will be strengthened as we've already looked at that.

The location I've chosen is in a small lobby area between bedrooms (so no noise above bedrooms), plus it's right next to a bathroom so drainage shouldn't be an issue.

Plumber is just wrong.

Regardless of where you put your tank, you will still have the same length of pipes outside the building, they are just either before or after the tank. So the only decision here is whether it is best to put the tank inside or outside, to which the correct answer is obvious.

As @-rick- says, you will run water through the pipes downstream of the tank more frequently, so it is even more important those pipes are inside the house.

As others have said, plumber is wrong.

We have decent insulation on our tank, and it's in a small, centrally located plant room that doesn't have any external walls. I'm still surprised at how fast the temperature drops. It would be waaaay worse outside.

2 hours ago, Bancroft said:

Looking for some advice on the placement of a DHW cylinder and buffer

Simple one first - buffer in the bin or better still don't buy in the first place.

DHW cylinder and heat pump spacing is something I know about as ours is quite a distance (22m each way). You get a great recorded CoP but use way more energy than you should - so much so that doing direct immersion heating doesn't cost you any more in energy.

1 hour ago, JpManning said:

primary pipes from the ASHP only lose heat while it's actually charging the tank, once or twice a day

The side effect is the cold slug of water cooling the cylinder at the start of the heating cycle, killing any thermocline. Cylinder is unusable from start of DHW reheat until the end.

So locate inside house and heat via immersion - simple.

6 minutes ago, JohnMo said:

Simple one first - buffer in the bin or better still don't buy in the first place.

Not sure we know enough about @Bancroft s plans to be certain of this. Need for a buffer depends somewhat on the overall system design. Blank sheet design should be able to eliminate but if other considerations are fixed it might be unavoidable.

6 minutes ago, JohnMo said:

DHW cylinder and heat pump spacing is something I know about as ours is quite a distance (22m each way). You get a great recorded CoP but use way more energy than you should - so much so that doing direct immersion heating doesn't cost you any more in energy.

The side effect is the cold slug of water cooling the cylinder at the start of the heating cycle, killing any thermocline. Cylinder is unusable from start of DHW reheat until the end.

So locate inside house and heat via immersion - simple.

@JohnMo always writes so authoritatively. I think it's always worth pointing out that the best option might be different in different circumstances. This certainly works for John, and he has shown good evidence that it will work for a fairly wide range of people but it's not a universal solution and wouldn't pass MCS rules (though given the content of this thread I don't think this is an MCS install).

IMO tradespeople have been known to favour whatever implementation is simplest and the least work for themselves. Their objectives are different from yours.

Whether they are on a fixed quote or a day rate, may or may not, influence this.

Such detailed design normally gets resolved before build esp if using an architect.

(Think carefully about access for monitoring, adjustment, and servicing. Who wants to climb into the loft to tweak the tank thermostat up or down by a few degrees. (I guess remotely readable thermostats or even cameras are available?)

Not directly relevant but just to note our past experience with a GSHP and "outside" DHW. The GSHP/GSHP buffer and DHW were in a "plant room" at the side of the (unheated) barn 20m from the house - ideal for avoiding noise in the house. I bought in to the idea that heat pumps work best on continuous at a low setting which suited our lifestyle as we're at home nearly all of the time so don't have the traditional peaks and troughs suiting (eg) a gas system where you get it hot very quickly - usually hotter than you want, then let it cool down and go cold when you're not there or asleep. By contrast we ran heating and h/w 24/7/365 - house was always 21, water 45. I ran "zero loss" Rehau pipe from the plant room to the house for both Heating and HW - expensive option at £50/m. Our flange heater (immersion equivalent) was only ever run for for the disinfection cycle for 2 hours a week and was horribly expensive in comparison! I also did a h/w ring on a pump so that h/w was never more than about 1m from every tap/outlet. The pump controlled on a 5 degree delta. HW was 45 degrees which the GSHP easily managed. Once the return fell to 40 the pump switched on and got the circuit back up to 45. It all worked wonderfully well for over a decade and we were very happy with it. However I'm pretty certain it wasn't the cheapest design to run....but it wasn't ruinous either so for the benefits I didn't really care.

2 hours ago, JpManning said:

Central is better for the hot taps anyway, short runs to the bathroom and kitchen save more over the years than the primary pipe loss

Good advice and take note of where the cold water supply enters the building.

59 minutes ago, -rick- said:

Not sure we know enough about @Bancroft s plans to be certain of this. Need for a buffer depends somewhat on the overall system design

But is almost never needed even with difficult systems, if you want multi room zones and/or have micro bore and want efficiency go Adia Thermal. TCVs that don't close, auto balanced system, primary and secondary pump speed matched and separated via CCT, even has the diverter valve, electric and heat meter and controls the heat pump over modbus.

3 hours ago, JohnMo said:

You get a great recorded CoP but use way more energy than you should - so much so that doing direct immersion heating doesn't cost you any more in energy.

Assuming an ASHP COP of 3, that suggests two thirds of the energy generated is lost during transmission through pipework. I'm far from an expert, but that seems high.

Admittedly I have very good insulation due to the pipework mostly going through the middle of my passivhaus slab insulation, but my ASHP is about 9 m from the DHW tank and UFH manifold, and I don't have anything like this experience. The temperatures reached at the DHW tank and manifold for DHW and UFH respectively are same as the corresponding flow temperatures measured at the ASHP, suggesting relatively little heat is lost on the trip from the ASHP.

25 minutes ago, jack said:

Assuming an ASHP COP of 3, that suggests two thirds of the energy generated is lost during transmission through pipework

We weren't particularly loosing temp along pipework, as all very well insulated. In fact the same pipework has been utilised for MVHR heating coil now, and today it was no warmer than 3 degs outside (about 12m of pipe outside), looking at temp difference at heat pump end and the end of the piping which once met the cylinder there was 0.1 degs temp difference.

We heated and still do twice per day, same number of people in house, same shower usage. We regularly got high 3s low 4s doing DHW. We used 4kWh of electric nearly every day. This all being measured via heat and electric meters

Doing immersion we use about the same electric.

I say ignore the CoP and actually look at the energy used to heat the water and test if it makes sense, I did and it didn't.

1 hour ago, JohnMo said:

But is almost never needed even with difficult systems, if you want multi room zones and/or have micro bore and want efficiency go Adia Thermal. TCVs that don't close, auto balanced system, primary and secondary pump speed matched and separated via CCT, even has the diverter valve, electric and heat meter and controls the heat pump over modbus.

That Adia Thermal thing is interesting but is somewhat besides the point. Different people have different priorities and restrictions. Limitations in budget, locally available installers, MCS rules, manufacturer warranty requirements, etc, may force some decisions. Maybe they aren't the absolute most efficient solutions but they may be the right solution for the situation.

Sorry to digress although slightly relevant. How noisy is a DHW cylinder when recharging? Ours ATM seems to be pretty noisy but we have a gas system boiler in current house. I am assuming they are a lot less noisy (silent?) if charged by immersion? What about if charged with ASHP?

Edited by Oz07

9 hours ago, JohnMo said:

Doing immersion we use about the same electric.

That's a surprising outcome given the numbers above. Minimal transmission heat loss and good ASHP COP should leave you using three or four times the energy using immersion.

Any theory why this is the case? I can't recall whether COP numbers account for defrost cycles, but even if they don't, they shouldn't be enough to neutralise the three or four times efficiency gain of the ASHP.

51 minutes ago, jack said:

That's a surprising outcome given the numbers above. Minimal transmission heat loss and good ASHP COP should leave you using three or four times the energy using immersion.

Any theory why this is the case? I can't recall whether COP numbers account for defrost cycles, but even if they don't, they shouldn't be enough to neutralise the three or four times efficiency gain of the ASHP.

Certainly were no defrosts, I do know that.

Did look at some systems on openenergymonitor and systems that heat once per day and others that heat twice. Those that heat twice can be using close to 3kWh of electric a day, those heating once around 1.5kWh. With a few doing half those numbers. But no idea cylinder size or water usage.

2 hours ago, Oz07 said:

Sorry to digress although slightly relevant. How noisy is a DHW cylinder when recharging? Ours ATM seems to be pretty noisy but we have a gas system boiler in current house. I am assuming they are a lot less noisy (silent?) if charged by immersion? What about if charged with ASHP?

It may depend on size of heat pump, when charged via a 6kW heat pump ours was silent. Change in ASHP size will alter flow rate, hence noise - maybe.

By immersion silent, but a previous cylinder we had was very noisy with immersion heater, really noticeable when on a PV diverter.

58 minutes ago, JohnMo said:

It may depend on size of heat pump, when charged via a 6kW heat pump ours was silent. Change in ASHP size will alter flow rate, hence noise - maybe.

By immersion silent, but a previous cylinder we had was very noisy with immersion heater, really noticeable when on a PV diverter.

Hmmm making me think about location next to bedroom. Strange that immersion would be noisy. What is making the noise the flow of water?

1 minute ago, Oz07 said:

What is making the noise the flow of water?

Sounded similar to a noisy kettle

2 hours ago, jack said:

Any theory why this is the case? I can't recall whether COP numbers account for defrost cycles, but even if they don't, they shouldn't be enough to neutralise the three or four times efficiency gain of the ASHP.

Each charge has a fixed + a variable component. My guess is that in John's case the spin up cost dominates the picture. Maybe reducing to one charge a day might make a difference (though John being the experimenter he is I assume he tried that).

John does have a more extreme case due to the long piping. The larger slug of water that has to be heated from CH/idle temps to start the charge and the wasted heat in the pipes at the end of the charge (might not be wasted if it's directed in the CH/UFH at the end in winter but might not count towards space heating in the monitoring). Actually, hang on, John, you said this 22m run was now being used to supply your MHVR coil. Does that mean you have two separate 22m runs from the heatpump to the house? Rather than one with valves inside the house?

With such a long run, I wonder whether it would make much difference to have a bypass valve near the DHW tank so that at the start of the cycle the water flow isn't directed into the tank but the heatpump just circulates the water in the pipe work until the temp matches the existing tank temp. Then the bypass closes at the tank is heated. Would save losing energy from the tank at the start of charge.

45 minutes ago, JohnMo said:

Sounded similar to a noisy kettle

Would you have near a bedroom or not? I suppose could acoustically insulate more than elsewhere

41 minutes ago, -rick- said:

Actually, hang on, John, you said this 22m run was now being used to supply your MHVR coil. Does that mean you have two separate 22m runs from the heatpump to the house? Rather than one with valves inside the house?

22m from heat pump to cylinder with UFH more or less at the mid point as piping enters house.

piping to MVHR now just simply tees off the main supply to UFH (used to have a diverter valve)

44 minutes ago, -rick- said:

With such a long run, I wonder whether it would make much difference to have a bypass valve near the DHW tank so that at the start of the cycle the water flow isn't directed into the tank but the heatpump just circulates the water in the pipe work until the temp matches the existing tank temp. Then the bypass closes at the tank is heated. Would save losing energy from the tank at the start of charge.

Did try that, didn't really make much difference to total energy used, did help mitigate some of the whole cylinder mixing at the start of the heating cycle.

54 minutes ago, -rick- said:

Each charge has a fixed + a variable component. My guess is that in John's case the spin up cost dominates the picture. Maybe reducing to one charge a day might make a difference (though John being the experimenter he is I assume he tried that).

Doesn't COP take into account spin-up costs?

And if so, how is it possible for an immersion to use the same amount of energy for the same result as an ASHP with a measured COP of ~4?

Create an account or sign in to comment

Account

Navigation

Search

Search

Configure browser push notifications

Chrome (Android)
  1. Tap the lock icon next to the address bar.
  2. Tap Permissions → Notifications.
  3. Adjust your preference.
Chrome (Desktop)
  1. Click the padlock icon in the address bar.
  2. Select Site settings.
  3. Find Notifications and adjust your preference.