HIGH FLOW RATE INSTANT TANKLESS ELECTRIC SHOWER

Creating a high flow rate tankless electric shower - by using wASTE WATER HEAT RECOVERY

Why?

Firstly, the justification for what you'll soon discover is a convoluted plan: I love powerful showers. And I really, really hate weak showers.

Calculating Energy Requirements

Usually there's two main ways to make a powerful shower: Either from a gas-powered combi-boiler, or from an electric or gas-heated water tank. Let's work out why these are the main options...

The amount of energy required to heat water for a shower is fully predictable, using a simple formula:

Q = mcΔT

where Q is the amount of heat energy in joules (J),

m is the mass of the water being heater in kg,

c is the the specific heat capacity of the water in J/kgK, for water this is 4184, and

ΔT is the change in temperature in °C/K.

The main thing we need to work out to begin with is the change in temperature. Here in Edinburgh, Scotland, the cold water temperature varies across the seasons, generally between around 5 and 10°C. And a good target temperature for a shower is 40°C. So this gives a ΔT of between 30 and 35°C. 

If we consider, to begin with, a single litre of water (which weighs 1 kg) being heated from 10 to 40°C:

Q = 1 x 4184 x 30 = 125 520 Joules, so this is the amount of energy required to heat one litre of water from 10°C to 40°C.

But we need to consider a flow rate, rather than a single volume of water. For this, we need to add a component of time. 

So to begin with, let's assume we heated this water over the course of one minute, giving us a flow rate of 1 litre per minute (lpm). So that's 125 520 Joules per minute to heat this water.

Now let's consider power, since this is the measurement used for showers, which is energy over time. The standard unit for power is the Watt (W) which is equal to Joules per second (J / s). 

For our 1 lpm flow rate we can now calculate Power = 125 520 Joules / 60 Seconds = 2 092 Watts = 2.09 kW

So that can be extrapolated a bit: For every 1 lpm flow we want, we need to provide a power of 2.09 kW.

Let's go back to where I started, I like powerful showers. And for me, a powerful shower means around 10+ lpm. So approximately 21 kW of power is required for this. If we look at the Screwfix website we can see the availability of showers:

The maximum power of shower available is 10.8 kW, clearly very far away from the required 21 kW. The reason for this limitation of shower power availability is twofold:

1. Capacity. A normal house in Scotland (and UK) has a supply and main fuse of 100 A (sometimes 60 or 80 A, but usually 100 A). The fuse protects the consumer from drawing more current than the incoming line and grid can safely supply. 100 A in power, assuming 240 V, is 24 kW. So drawing 21 kW for a single shower would take the household consumption close to the capacity, limiting the use of anything else in the house without risking blowing the main fuse.

2. Diversity. Because of this capacity issue, when an electrician installs an electric shower into a house, he must consider what's called 'diversity'. This is a calculation which works out the approximate total load of the house at present. This sounds simple, but is actually quite a complicated and non-scientific calculation which apportions current to each use point based on maximum theoretical current draw, as well as frequency of use etc. For example, a cooker is calculated as 10A, plus 30% of the remaining theoretical current draw, since it is very unlikely every heating element will be on simultaneously for any great length of time.

When calculating the diversity / maximum, the demand an electric shower contributes to the total cannot be reduced in the calculation, since it draws 100% of it's theoretical current for 100% of the time that it's on. This, along with the cable size requirements (the more current that an appliance draws, the larger the cable must be) limit the actual power numbers you see on the Screwfix website above.

Note - it is possible to have a supply to your house of >100A, by installing a 'three-phase' supply. This is basically three live wires, out of phase from each other by 120°. This could give you 300A, and is common on very large or commercial properties, where 100A may be insufficient. For example, I used to live in a flat in Aberdeen which was originally a police station. This had a three-phase supply, though only one was in use and metered.

So why didn't I just use the maximum available 10.8 kW shower shown above? Well, let's look at the data:

Flow rate is shown in lpm on the y-axis, assuming a 30°C temperature rise. These are some very terrible flow rates, and this graph is the reason for all of my complex plans which we'll soon discuss. Even with 10.8kW shower I'd only be able to achieve ~4.5 lpm in winter, hugely less than my target of 10 lpm, and a pretty poor shower.

What are the Alternatives?

Ok, so an electric shower as-is was out of the equation. So what are the normal alternatives?

1. Use a supply from my main house.

In my main house we have a 40kW combi boiler, so in theory I could have taken a feed from this to supply the micro house shower. However, there was some issues with this: Firstly, we use 100% of the 40kW for our morning showers, and we love it. I'm sorry mother nature, but this is one area of my life where I don't think green. There's nothing better than an embracing waterfall to warm up and get ready for the day. So we didn't want to impact this, by siphoning off half the flow rate. Since the micro house may sometimes have guests staying in it, I also didn't want to limit when they could shower based on what's happening in a different building. So using half the flow from this boiler was out of the question. But we'll come back to this boiler...

2. Use an electric heated tank.

An electric or gas heated water tank is a usual solution to all these issues. Since the tank can be heated low-and-slow, you can heat up a large volume of water without taking out a lot of your supply capacity. You can also then use a pump to achieve as high as flow rate as you like, based on the size of the tank, number of users and length of shower. However, I'm building a micro house, and one without any kind of void/loft, so this was out of the question unfortunately. 

3. Pre-heat the water supply (and add some more power...).

This is where I found my solution. Since I'm focussing on flow rate, there are only 2 means by which to impact this - power and initial water temperature. If I could find a way to raise this temperature, I could have substantial improvements to the calculated flow rate at a given power:

Here you can see the substantial impact that increasing the initial water temperature has on the realised flow rate. The black box is my 'target zone' of >10 lpm. You'll see now that a couple of extra shower powers have appeared - 12 kW and 14 kW - we'll come back to this.

I thought about a lotttttt of possible solutions to increase the incoming water temperature, and settled on two, one to implement from the start of the build, and another to implement further down the line:

SOLUTION 1

A waste water heat recovery system (WWHRS). These are devices which transfer some of the heat in the used shower drain water back to the incoming clean/cold water, prior to it being heated. There are various designs, but the one which I liked the look of best is shown below:

This is a RECOUP PIPE HEX RD ACTIVE. This consists of a double-walled copper drain, with the drain water being pumped from the shower waste upto the top of the stack and released to the central pipe in each tube, via a cyclone device which encourages the water to flow along the pipe wall. Cold water enters the outer pipe at the bottom. This then flows up, in the opposite direction to the drain water, before going on to a water heater.

This device works on a principle called 'counter-current heat exchange'. 

Interestingly this is the same effect your body uses in cold temperatures, and is the reason your toes and fingers get so cold! Basically, it is theoretically possible to entirely 'swap' the temperatures of two fluids which are run in opposite direction to each other, contacted through a thermally conductive surface, but not mixing, if you give enough contact/time/distance. This WWHRS can achieve about 50-65% efficiency at transferring the heat from the drain water to the incoming cold mains water.  

SOLUTION 2

A small stream of water, taken from the boiler in the main house. Enough to set off the boiler to heat, and to contribute a good amount to pre-heating, but not so much as to have a noticeable effect on the shower in the main house. 

In planning for this, when I dug the service trench to the micro house, I added a conduit with an insulated 10mm plastic water pipe. I chose 10mm to intentionally limit the maximum flow rate which would come through this pipe. That way there would never be any means by which to significantly disturb the flow rate of the main house shower, even in the event of a leak or error. 

The final plan

So let's have a look at the final plan:

Please excuse my terrible drawing skills. The bowties are lever ball valves. So, if we start at the top, what we have here is the hot and cold supplies coming in, supplying the workshop sink. The hot is then combined with the cold and flows either to the WWHRS or to the water heater directly. The water heater then heats this water and supplies a normal thermostatic mixer shower. Here, the relatively unknown hot water temperature will be brought down to ~40°C by mixing with some more cold water supply (I changed this layout later). When the WWHRS is installed, the hot supply from the main house will be permanantly closed off. However, it will still supply the workshop sink, to provide a lower-cost hot water source here, where it may be required to be used for long periods.

Missing from this diagram is the bathroom sink, which is supplied from the water heater, not the 10mm supply. I think that the sink flow rate will be low enough that the main house boiler will not activate, and therefore the hot supply will come from the water heater only. We'll see if this works out as I'm imagining. 

The next step in achieving this plan was to identify a water heater that can function in this set-up. This was relatively difficult. I found that most Mira-type showers had an inlet temperature limitation of around 25-30°C, which may have been somewhat limiting to my plan, or would force it to be a bit too finely balanced. I wanted something more tolerant than that. This is what I settled on, after discussions with the technical team at Stiebel Eltron (not sponsored, no relationship with company):

This water heater has a maximum inlet temperature of 55°C and no stated limitation on flow rate that can pass through it. It has a variable outlet temperature set-point, but (as I found out after install) it cannot limit flow itself, beyond a small capsule flow limiter which I immediately removed. So the temperature set-point is achieved by varying the applied heat/energy. The maximum energy applied can be set by moving a jumper within the unit, at either 12kW (50A) or 14.3 kW (60A). The effective current load can also be curtailed by limiting flow, such that the set-point temperature is achieved at a lower applied energy.

It was important to be able to control pressure within the system, to ensure that the slightly lower pressure hot supply (~0.5 Bar lower from flowing up to and through my gas boiler) would flow into the cold sufficiently. It was also important to have flow limitation, since the incoming ~6 Bar 25 mm supply (22 mm from micro house penetration to water heater) would supply about 80 lpm if left unlimited (data from testing). So I decided to add pressure control valves (PCVs) on the incoming hot and cold supplies, to allow me to balance and limit the flows.

The Initial Construction

This is what I ended up with after phase 1 of the plumbing (keep scrolling for phase 2 with the WWHRS:

The left image here can be mentally plopped on top of the right image. So, starting with the right image, at the bottom right, we have a 22 mm cold supply, and a 15 mm hot supply (converted from 10 mm plastic the other side of this wall). Both of these pass through non-return valves (NRVs) and PCVs. You can see a couple of pressure gauges which can be used to help balance these supplies. They are then combined into a pre-heated supply that feeds the water heater. You can see I also already have the pipework in place to easily add in a WWHRS if/when I do this. 

The Electrical Supply

You can see at the top of the water heater image that I used a pretty beefy supply cable - a 16mm2 twin and earth. This is the required cable diameter for a 60A load, when 'clipped direct'. You can also see a 63A isolator, quite a chunky thing. The supply cable comes from a 50A RCBO. This is less than the 60A maximum load of the water heater, but is the biggest Hager RCBO I could find. From my testing so far, this RCBO is ok with 60A running through it, at least for the 10-20 minute length of a shower. However, it may reduce the lifespan of it, since RCBOs get hot when at/near/above their rated current.

But What About that Diversity??

Ok so I've put a 50-60A current user on my 100A supply. That's a lot. And I think most electricians would probably not do this, because no diversity is afforded to electric showers, so this reduces the effective capacity of the remaining supply to only 40A. However, diversity is not an exact science, and a degree of flexibility is allowed, considering the prevailing circumstances. In my case, the building is primarily going to be used now and then by guests, and therefore it is an occasional load. And when it is used, the timing of use is approximately known. Anyway, at present, the only other large load in my house is a 32A Ohme Home Pro car charger. This has a CT clamp over the incoming supply, monitoring the total load of the house, so it shouldn't operate when we're near the maximum load. I also have CT clamps monitoring the micro house (I'll talk about these in another post), so in future I may smart-ify the shower, to ensure it cuts out or doesn't operate if there isn't capacity. I think this is the future of 'diversity'. We're going to have more and more circuit loading on our homes as heating transitions to electric, and electric cars prevail, so I think smartification of current loading is inevitable. 

Testing

I was finally able to test this set-up, three years after initial design, in October 2024. So let's talk about my first test results. This is just a rough summary for now, I'll do some more comprehensive testing in future and come back and update this section with graphs etc.

Firstly I looked into the behaviour of the water heater itself. I ran through various flow rates and observed the final water temperature, and the power consumption of the unit, measured via a CT clamp onto an app on my phone. When set at 60°C, the heater can very consistently provide that, with a decreased power use at lower flow rates. At higher flow rates the power maxes out, and the temperature of the water drops off. So, as expected. But what about with the pre-heating?

With the hot water from my boiler tied in, I was able to flow ~8.8 lpm at 53.2°C. Calculating for the theoretical flow at 40°C, once cold water is mixed in, this gives a flow rate of 12.6 lpm. So it worked! However, this was only the first part...

Installing the WWHRS

I was finally able to install the WWHRS in April 2025, and the final, complete installation is shown below:

So, there's a lot there... The system I got from Recoup has a pump that's positioned inside the floor of the micro house. It's probably impossible for you to follow the pipes, so I've drawn up a diagram to help:

My final installation is a fair bit more complex than the 'plumbtastic' plan. This is primarily for my own interest and testing purposes. A WWHRS can be installed in a number of 'systems' / flow regimes:

System A (left) provides pre-heated water to both the cold supply to the shower mixer, and to the water heater / hot supply. System B supplies pre-heated water only to the cold supply to the shower mixer, and cold water to the water heater. System C provides pre-heated water to the water heater only. 

The three different systems provide three different heat recovery efficiences:

So system A is the manufacturer recommended method of installation, and theoretically has the best performance. But I'm a scientist and engineer, so this table made me want to do some testing and verification! I also wanted to keep my options open, incase there was a problem with one or more of these methods. Specifically,  I was somewhat worried that there may be a 'hunting' effect in temperature. So, I also wanted to make sure I wasn't putting all my eggs in one basket.

So let's test it!

First Test with the WWHRS

My first test was nerve wracking. This had been planned for years, and I really had no idea whether my design would actually work. For the first testing, I installed my bar shower with thermostatic mixer temporarily in the bathroom, and just stuffed the shower hose down the drain. This is not ideal, since it does not give me the cooling effect of water flowing over a person/over the floor, but was fine for some initial tests. I hadn't installed the membrane on my wet room floor yet, so couldn't run the shower properly.

The shower I went for is the Hansgrohe Crometta S Showerpipe 240 . I was tempted by concealed units, but thought I'd keep it simple for my first ever bathroom install and just get a bar mixer. I chose a good brand, since I knew I would be testing that thermostatic cartridge!

For the initial test I used 'method A'. Here's the results:

Cold water temperature = 10°C

Shower water temperature = 40°C

Pre-heated water temperature = 26°C

Recorded flow rate = 15 lpm @ 40°C

The temperatures were mostly measured using a thermal camera, and a probe for the shower temperature. The flow rate was measured by timing the fill of a 2 litre jug. A flow cup was also used for less reliable checks.

Side Note About the EU / Efficiency

If you look at the manual for the water heater I used, you'll see this:

Queue a lengthy back and forth email exchange with Stiebel Eltron's technical guy. An electric heater is a resistive load, which should be 100% efficient, neglecting the small amount of losses in heat within the unit itself. So I could not understand why this had such a terrible efficiency...

I eventually found the answer, in amongst some EU directives:

"(13) ‘conversion coefficient’ (CC) means a coefficient reflecting the estimated 40 % average EU generation efficiency referred to in Directive 2012/27/EU of the European Parliament and of the Council (1); the value of the conversion coefficient is CC = 2,5;"

In simple words - it's a pile of stupid nonsense. It's trying to say that the heater is 39% efficient because the supplied energy was generated at an average 40% efficiency, across the EU. And so the whole process of converting gas and uranium and wind into heat within my water heater was 39% efficient. I cannot even fathom the ridiculousness of that logic, but that's the reason nonetheless. The reality is that the heater is ~99% efficient at converting the supplied electrical energy into heat energy in the water that passes through it. I've always been pro-EU, but inexplicably stupid directives like this turn me momentarily brexiteer...