I am wanting to hook up a solar system to my house but don't want to feed back into the grid (for a number of reasons, including legislative bureaucracy and having to change my electricity company). Notwithstanding it may make sense for me to spec my solar system so it produces more power then I need (future proofing, handling the differences between summer and winter collection)

When I explored this years ago, I recall I was advised to dump the excess power into a hot water cylinder or other arbitrary load. I do not have a suitable hot water cylinder or similar wasteful load - although I guess I could always rig one up if I absolutely had to.

I have discovered a hybrid converter where, when asked, the installer claimed "The solar inverter will only take energy that is required from the solar..."

Is this correct [ that MPPT controllers can do this ], and if so do I need to worry about what happens to the excess power and any effects it could have on the panels?

  • Depending on where you live (esp. whether you need heating) you definitely should consider a thermic use of solar power. Yes, you need a buffer of about 1 cubic meter, but the reduction in heating and hot water cost is significant. Once the sun is out, the panels give you 50 ° C hot water even on Christmas at 0 ° C air temp. As to your actual question... sorry, I have no idea.
    – Erik
    Jan 8, 2021 at 7:53
  • @Erik As it happens I have a large hothouse on my property - so I guess I am already making thermic use of solar power! I'm not sure what you mean by a buffer of 1 cubic meter, but I know as a very crude approximation the raw power of the sun when overhead is 1kwh/m^3 Of-course, when the sun is up I don't need much heating.
    – davidgo
    Jan 8, 2021 at 8:06
  • The 1 cubic meter is the approx. volume of a buffer tank you'd need to heat a "normal" family home (ground floor, top floor, about 3 bedrooms).
    – Erik
    Jan 8, 2021 at 8:10
  • You say you explored this years ago. Is that also when you decided not to work with the utility? This is a rapidly changing space, so you may find that the utility has improved their offerings.
    – LShaver
    Jan 8, 2021 at 17:08
  • @sspring this is really about the solar panel - you don't want to overcharge any kind of battery. I don't think an aircon is a practical arbitrary load as you will stuff up the compressor in no time if you treat it that way. On that matter, do you know people who actually use a chest freezer as an on-demand load?
    – davidgo
    Jan 10, 2021 at 3:13

2 Answers 2


the installer claimed "The solar inverter will only take energy that is required from the solar..."

This is how all solar inverters work. You can't not get this "feature", because the system would blow up the first time it was turned on if it didn't work this way. Think about it in the context of no load - you power up the inverter but for whatever reason the circuit breaker pops or the heater turns off... either the inverter generates no output power, or the output voltage rises until something starts arcing.

Many solar inverters also have output limiting, so they can be set up in places where the grid operator has rules limiting how much each system can feed into the grid (commonly in Australia it's 5kW per phase, looks like in NZ it's often 10kW per system). Those inverters (and some others) can be set for zero feed in. Talk to your installer, they will almost certainly have a system that can do this.

BUT... Meridian, and every other supplier in NZ, don't charge extra to allow you to feed back. So the question is really: would you rather be paid something for excess power you have available, or nothing? The cost to you of turning that power into grid-comaptible power is almost unmeasurably low (it's the marginal wear and tear from extra power through the inverter, likely to be (much) less than 0.1c/kWh.

(edit in the rest...)

When I ran the numbers on a few different options for my house (in Sydney, but the basics are the same), most things never paid for themselves. I use off peak electricity for the resistive hot water, for example, so the maximum possible saving there is the 30c-40c/day I pay now. Even insulating that tank with free rockwool and a couple of $5 plastic drop cloths took more than a year to pay off. That's when I was selling electricity for 10c/kWh and paying 12c/kWh for off peak.

Using a programmable relay to switch your entire output to a heater of some sort is the cheapest option. You need an electrician to do this, but sadly you also need an off-grid inverter because it's not going to be connected to the grid when it's switched over. So the $200 relay turns into an extra $1000 or so for the inverter.

Or you can get a controlled load of some sort and match that to your excess solar (easier than it sounds, a clamp meter to detect current on the street feed wire in your meter panel may not even need an electrician to install it). The trouble is with the load. A resistive heater is the obvious option, since that won't care that the supply goes from 0% to 100% in a couple of seconds when the sun pops out from behind a cloud, then back to zero a minute later with another cloud. But... those things aren't off the shelf parts. You can do some of it, but you'll need an electrician to install the heater and controller and that will cost $1000 or more for parts and labour. So now you need to save $1000 worth of 8c/kWh electricity to break even on that, assuming the actual heater etc is free.

I couldn't come up with a way to make it pay with a grid-attached system. Off grid it's easy... the MPPT controller has a programmable relay, that turns on a 48V hot water heater when the battery is charged... voila, free hot water (in the summer, anyway)


Yes, that is how solar controllers work normally.

Now, a grid-tie charge controller+inverter has one job: Shove as much power as possible back onto the grid while the grid is up, and shut off entirely (per UL 1741) when the grid is down.

But plain (DC) charge controllers coupled to a battery, need to throttle their output up and down so they charge the battery at appropriate rates and "tail off" the finishing charge as the battery nears 100%. This varies "on the fly" based on whichever DC loads are being powered.

It is no problem for the controller to reduce its draw to less than the solar panels are capable of putting out. If the controller reduces its draw, solar voltage will rise. But it will not rise infinitely; it will stop at the panel's float voltage (commonly 19V for a "12V" panel). At that point, instead of making power, the solar panel will just sit there getting warm, but that does not harm - it'll get no warmer than any other black thing sitting in direct sun.


Better-featured DC solar charge controllers have a third set of terminals called "DUMP". The charge controller provides DC power to these terminals when the solar is making more than can be put into battery + direct loads.

A DC charge controller with "Dump" make the water heater exercise rather easy: simply route the "dump" power to the heater.

The "dump" power could be fed through a DC-AC inverter to make a correct voltage for the heater (which needs to be AC because switching 120/230V DC is very difficult, and AC-rated heaters are not rated to switch DC). This could then pick up a DPDT contactor which would switch the heater from utility to solar when solar is present. However, here's another thought.

Many water heaters have 2 heating elements, and only one is used at a time based on circumstance. The water heater could be "hacked" to use one heating element for utility AC, and the other element replaced with a DC element of your system voltage. This would reduce overall heater performance (it has 2 heaters for a reason) but would allow you to capture the excess energy.

Or water could be recirculated between the water heater's bottom drain valve and a "tee" off the cold water inlet, using a small pump and an inline heater. This would need to be well insulated or it would defeat the purpose.

Temperature management

You have to plan temperature carefully. You cannot have water sitting in a heater tank much below 60C (140F) or it will breed legionella and other bacteria. (and if this is news to you, it's because it's fairly new science). This already requires thermostatic faucets. So you can push temperature higher, e.g. up into restaurant dishwashing temperatures, but you can't go near 100C or you'll over-pressurize the heater and make the safety valve dump water all over your floor (or worse: fail to do so, causing a BLEVE!)

Your best bet might be a timer -- which switches the water heater from utility to solar in the morning just before you do your morning bathing (so that depletes its hot water and gives the dump energy something to heat)... and switches it back at dusk or whenever the favorable utility rates kick in.

It takes longer than 12 hours for bacterial growth to flare up significantly in a water heater, so as long as it's getting heated to ~60C every night, that will definitely kill the bacteria.

  • (BLEVE = explosion)
    – user253751
    Jan 27, 2021 at 22:14
  • I thought BLEVE was only an issue if the compressed gas was itself flammable?
    – LShaver
    Jan 28, 2021 at 19:06
  • 2
    @LShaver Oh no, that's only a bonus. If the liquid is heated to above boiling@ambient, and then suddenly the vessel breaches and pressure falls to ambient, then some of the liquid boils to vapor instantly to consume the excess temperature. 1 BTU = 1 pound of water heated 1 degree. So if you have 200lbs of water at 5 degrees F above boiling@ambient, that's 1000 BTUs of excess energy. 1000 BTU is the enthalpy of 1 pound of water, so it will instantly turn 1 pound of water into steam. 1 pound is a lot of steam. Jan 28, 2021 at 22:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.