Wednesday, 11 July 2012

Water Heating Experiment.

Pete has sent in the following: March 2012
I've read with interest various experiences regarding heating water with solar PV. The majority seem to have given up without posting clear conclusions. The main problem appears being able to utilize the power generated by the PV’s to power an immersion heater which is normally rated at 3kw, without drawing any shortfall in power from the grid. Obviously, even with a largish 4kw system, there is unlikely to be significant opportunity to generate this level of power and certainly not consistently for an extended period.

Notwithstanding the above I can see a clear benefit from using solar pv to heat an immersion, if electricity or oil is the main and only energy available. The reason being that it would be necessary to use an immersion in any event, so any means to reduce the draw from the grid is beneficial. The benefit (financially) becomes more uncertain however, if there is a gas supply. There would however still be a benefit from reducing total energy draw and being green.

My comments below really relate to situations where a gas supply is present, rather than electricity or oil being the only energy available. (As I consider it’s a bit of a no brainer to use pv solar to power an immersion if you were going to have to power one anyway)

In order to maximise potential gain, it is possible to fit a timer for mid day use to switch on an immersion for a couple of hours when maximum gain (over 3kw) is anticipated. However, if it clouds over for periods, you end up drawing excess from the grid. That is, you will be paying for the difference in energy being generated by the panels and the 3kw draw from the immersion. This could be a lot, and if like us you have a gas boiler then this is just nonsensical.

To increase the potential to utilize the generated power, the immersion heater itself could be replaced with a 1kw unit. There is still an element of hit and miss though, and a timer is still required to limit the draw to a period of say 11am to 4pm (summer). A 1kw immersion costs about £50 and you will probably need a plumber to fit it, unless you know what you are doing, you obviously need to drain the cylinder. There is still the problem of potentially importing energy from the grid though when clouds reduce generation below 1kw. (more realistically below 1.3kw,   as there will be background use going on).

An alternative is to keep the 3kw immersion in situ but fit an 110v transformer in the line of the immersion circuit. This will automatically reduce the draw to about 700w. A timer set for between 10am and 4pm (in summer) should have quite a high hit rate as the draw will be limited to 700w. While in theory a better solution, there are still going to be dull or cloudy days when 1kw (more realistically 1.2kw due to background uses) will not be produced and you will either have to turn the immersion off, or pay for the excess drawn from the grid.

A further refined option is to use the transformer (reducing the draw to 700w) and fit a relay switch between the PV generation meter and the consumer box. The purpose of the relay is that as soon as 1kw is being generated by the panels, a trip switch opens onto the immersion circuit allowing the immersion to draw its 700w. Anything over 700w goes back into the circuit as normal to be used by other appliances presently in use or back into the grid if not used. If the generation drops below 1kw due to cloud cover or being towards the end of the day,  then the switch automatically closes and the immersion stops drawing energy. Thus the immersion is prevented from drawing energy not being generated by the solar panels. If the immersion heats the water to the desired temperature, the thermostat will stop energy draw as normal.

This is what I have fitted to my system and am currently attempting to collect data to judge the effectiveness and payback. Its only been fitted for a month, so its early days.

I sourced the necessary switches, and my solar panel fitters installed the bits for me, when they commissioned my system. The system is very easy to fit and having looked at the instructions I could have done it myself. However, building regs prevented me from doing so. Thus anyone thinking of installing this kit should employ an electrician for an hours work (at most).

The thing which interests me about this project is whether it will be possible to heat a tank of water by solar PV, or at least pre heat, and thus cut or reduce the cost of it being heated by my gas boiler. Obviously, if the water is heated to the required temperature then the boiler will not have to fire up. If however, the required temperature is not reached, the boiler will have to fire up but for a shorter period as the water should have been pre-heated, thus still saving gas.

The goal is to establish whether it is possible to operate this system in a way which reduces our gas bill by consistently using energy in the form of a down rated immersion heater used over a large part of the days generation period. That is energy which would otherwise largely go back into the grid as we have days when the house is unoccupied during the day so heating water for the evening baths appears to be a solid option.

Further modifications include a timer switch which can cut off the immersion for a set period in order to use washing machine etc which could also be on timers.

As a household, our main energy draw between April and October is from heating our water. We have three small children and thus there are frequent baths and showers. We have a large 300ltr cylinder and it is hoped that the immersion modification will at the very least help pre-heat the water and thus reduce our gas use during this period. We fully appreciate that the potential of this system during November to March will be limited as we are not expecting too many occasions of consistent extended generation periods in excess of 1kw. There will be some however, and every little bit will help. (actually the last 5 days have been spectacular)

I have kept good records of energy generation over the past year or so, and can thus join an esteemed group of anoraks. What this does mean is that I have a reasonable understanding of useage. There is a fly in the ointment however. In September, we fitted a new condenser boiler and a 300ltre tank as part of an extension project. Thus my data for last year will not be directly comparable as I have a more efficient boiler heating a larger volume of water. I will do my best to make a judgement however. What I have noticed however, is that the water temperature is hotter, as I’ve set the immersion thermostat quite high. This means that we do not now have the boiler set to come on in the early hours for morning showers. In the evening, at present the boiler is set to come on for 50mins. I’m not sure how much of this time the boiler fires up for, as I’m not home at the time. This does however give us enough water for evening baths and as mentioned above morning showers.

Given that I already had a 110v transformer (probably about £50 from screw fix or the like); the outlay has not been massive. The switching gear has cost me £180. Not an insignificant sum of money I know, however, it’s now paid for and hopefully I can look forward to reducing my bills over an extended period, plus I will be beneficially utilising a larger proportion of the energy which I am producing myself.

It’s very early days and to be honest an anoraks experiment, however, I will provide an update as and when.

If you wish to reply, or add a post of your own please email: 


  1. Given the summer most of of the UK is now experiencing I consider a relay switch an essential component of the system rather than a refinement.

    For those interested I've recently uncovered a commercial offering to what's being discussed here:

  2. I'm doing a similar thing but with off grid solar. A sensor measures the strength of the sun, estimates the main PV array power potential and takes into account the charge state of the lithium battery bank before deciding to turn on of off the immersion heater. It uses a 110V transformer but provides two power levels by switching in or out a diode on the transformer output. This gives full cycles for 650W power or half-wave rectified cycles for about 300W power. The solar sensor means that the system tracks the weather with 15 second resolution. It also tracks other loads in the house and if the kettle or washing machine goes on, the immersion heater will turn off automatically if the load is too great for the solar power available.

    I'm also working on a prototype of a phase angle switching controller for grid tied solar use that can switch the heater in fine steps from nearly nothing to full power on any single phase heater up to 6kw. Like a lamp dimmer on steroids.

    The switched transformer does quite well though and as we have a small tank, on good days I get a 100 litres of 50-60C hot water and we need use no gas for the central heating boiler.

  3. Hi drambo,

    I have also tested out the Intelligent Solar Immersion Switch mark II ( ) and it works brilliantly. It monitor the pv production and the house hold load with a single sensor clipped on to the grid lines !! That makes it much easier to install. Also it uses the existing immersion wirings from the consumer unit.

    The unit turns on the immersion when enough surplus power is produced. It will turn it off when the house hold load increases - ex. if a kettle is turned on.

    The unit is also affordable and clean looking.

    Thy also do a 3kW compact power reducer, which lowers a 3kW immersion to a 1.5 kW. Probably better clean option than a site transformer.

    Had a chat with technical support last day and looks like a new version with proportional controller is on it's way.

    I recommend this unit to any one having Solar PV installed and want to use the surplus power generated.