18 hours ago, Denns said:

That’s very nice. 14 yrs service is amazing to be honest.

Mine is now into it's 14th year. A Kwikot. And they still have spares and support for it.

The other problem with hooking up a geyser directly to solar panels is the same problem you have with other types of solar geyser. What if the sun doesn't shine?

I had a solar geyser at my previous home. It worked well most of the time, but it had an electrical (230V AC) backup in place, controlled by a timer and a thermostat, for days when the weather wasn't so sunny.

The heat pump seems like a good way to go. It doesn't require sunny weather. It does require electricty but not so much as an element does. As an example, I run mine at 6:00 in the mornings because I like an early morning shower. Days when I am going to the office (I mostly work from home) I start it at 5:00. Even starting at 06:00 in the summer it mostly runs from battery and without the sun shining on it. It costs me very little, it always gets over 50 degrees (thermostat is set to 57 and I can't set it higher than 60) despite weather or (when it was a thing) load shedding.

22 minutes ago, Bobster. said:

The other problem with hooking up a geyser directly to solar panels is the same problem you have with other types of solar geyser. What if the sun doesn't shine?

I had a solar geyser at my previous home. It worked well most of the time, but it had an electrical (230V AC) backup in place, controlled by a timer and a thermostat, for days when the weather wasn't so sunny.

The heat pump seems like a good way to go. It doesn't require sunny weather. It does require electricty but not so much as an element does. As an example, I run mine at 6:00 in the mornings because I like an early morning shower. Days when I am going to the office (I mostly work from home) I start it at 5:00. Even starting at 06:00 in the summer it mostly runs from battery and without the sun shining on it. It costs me very little, it always gets over 50 degrees (thermostat is set to 57 and I can't set it higher than 60) despite weather or (when it was a thing) load shedding.

For my case, generator. Assuming a 6KVA one and running it at 80% of rated output for 3 hrs a day 20 days a year. Comes to about 2520 rand in fuel cost. 3 hrs of runtime is plenty to heat my geyser and charge up my batteries.

That is 2 months of fixed charges under CP postpaid or a year of current fixed charges on prepaid. That is why I have no interest being connected to the grid. With a gas generator the running cost is even less. No ROI with the grid anymore for my case.

2 hours ago, Denns said:

Such modifications may work for years without issue, but all it takes is one incident to create a big mess for oneself.

I just went with a geyser controller that comes with certification, etc. It was only 2.7k. Allowed me to connect the 4 panels to the geyser element since it outputs a modified sine wave to the thermostat/element.

A properly rated DC contactor can even hit 1k in cost. It is just not worth it.

@Denns which controller are you using?

14 minutes ago, Leb7 said:

@Denns which controller are you using?

Geysertech PI2R micro. Hooked up to 4x340W panels. Total installation was 9k.

Then heat pumps - not great. For grey days I planned on having an instant gas geyser downstream that would only heat the already warm water if it wasnt hot enough - 3 years now, never needed - I do have the solar electric thermostat at max (sun is free). All electricity is dangerous and should be treated with respect and caution; however “AC currents and voltages are three to five times more dangerous than DC having the same level of voltage.

18 minutes ago, tony Lampard said:

All I want to add to this post. Using an AC relay to switch the not so dangerous arcs of DC is a mayor myth.The same can be said about AC MCB's in DC circuits. Even at low currents and high DC voltages that can sustain the arc during opening. That is the reason why one has to follow SANS regulations even if the element will be happy to work on AC or DC.

Secondly I have over 14 yrs of owning one found 2 statements on heat pums and normally from those that had a bad installation or never had one. The excessive noise and high cost of maintenance. Funny that with no higher cost than an aircon which nobody complains about is quite strange. Yes mine is now running for 14 yrs and apart from blowing clean I only had the feed pump replaced at a R1200 cost and this was due to another error by the installer by not fitting a strainer to prevent grit from entering the pump. If one opens a normal geyser without strainer one will see how much grit accumulates in a geyser in say 5 yrs.

7 hours ago, tony Lampard said:

"the not so dangerous arcs of DC is a mayor myth" - all that will happen if the contacts arc closed is the geyser will remain on. The absence of thermostat control is a feature of thermal solar water heaters anyway (they boil) - so what's the biggy? I have double break on both + and -; through a Normally Open (NO) and back through NO with red; same again with black - 3 years down the line there is no sign of arcing. Very well earthed. If your bathing preference is when the sun is down, that pretty big switch up in the sky also comes into play. This is open source information played forward, it is not a sales pitch; my thanks go to very competent people for giving me free advice (the double break through the contactor advice came from an electrical engineer friend) while doing the thorough research following picking up the conceptual idea off the internet and a proof of concept trial connecting a geyser element to a solar panel in the garden. 4 x 450watt panels to feed the 3kw element on a Kwikot pressure geyser was the biggest cost. If Kwikot jump on this band wagon and fit two elements each with its own thermostat to their geysers you could have one of the elements fed from the conventional AC with a time switch for Demand Side Management (DSM) and the thermostat set low (grey days fail-safe) and the other off DC with the thermostat set high; they could also supply the control box. I would like them to use more substantial electrical connections on their elements, they have a very flimsy configuration - that design is very micky mouse.

Whether it works fine or not is irrelevant for most people in my opinion. The question is will someone like @TaliaB who I think is a PV installer give a COC for this?

it’s all for insurance. If your house burns down then insurance won’t pay. It doesn’t matter if it worked for 10 years.

My geyser controller doesn’t output AC. It outputs a pulsed DC straight from the panels. It has all the certifications and is approved for use in homes. Guess what happens if my house burns down? They can’t blame my controller etc.

The 2.7k a spent on the controller is the only real additional cost my system has over yours and even if we don’t deduct the contactor, is cheaping out on the controller worth not having peace of mind for insurance?

The people that designed the controller decided to make it a pulsed DC wave and went through all the trouble to get the unit certified. If all was fine and dandy with connecting panels to an element straight then they would not have bothered.

As for heat pumps I think they are great. If I had the money I would have done it. Using 1kw to provide 3kw of heating power is not a bad deal. And @Scorp007 and @Bobster. personal experiences have sold me on it.

11 hours ago, tony Lampard said:

The absence of thermostat control is a feature of thermal solar water heaters anyway (they boil) - so what's the biggy?

That statement is technically wrong, dangerously misleading, and shows a fundamental misunderstanding of both electrical and thermal safety principles.

Any installer or advisor suggesting that “lack of thermostat control is fine” would be liable under both electrical and plumbing safety regulations if someone were injured or property were damaged.

It would also void any warranty and fail any inspection for SANS or NRS compliance.

The Compulsory Specification (VC 9006) states that fixed electrical storage water heaters “shall comply with the requirements of SANS 151” for component safety, including thermostats, electrical safety, connections etc.

In documentation about geyser construction, it is asserted that geysers are supplied (and must be maintained) with a thermostat, a T&P (temperature & pressure) safety valve, and that the thermostat must “comply with SANS 181: Thermostats for electric storage water heaters.”

So the “absence of thermostat control” is in direct conflict with the requirement that heaters comply with SANS 151 / the Compulsory Specification, which inherently demands safety devices.

Geysers are not a kettle it doesn't, must not boil hence the reason a Kwikot geyser thermostat maximum temperature setpoint is 70°C intelligent setting would be between 50°C~60°C

Edited October 12, 2025Oct 12 by TaliaB

14 hours ago, TaliaB said:

  On 2025/10/12 at 7:15 AM, tony Lampard said:

The absence of thermostat control is a feature of thermal solar water heaters anyway (they boil) - so what's the biggy?

That statement is technically wrong, dangerously misleading, and shows a fundamental misunderstanding of both electrical and thermal safety principles.

Any installer or advisor suggesting that “lack of thermostat control is fine” would be liable under both electrical and plumbing safety regulations if someone were injured or property were damaged.

I think the reference is to a certain type of geyser that heats water using heat from the sun and that heating does not use electricity in any way.

I used to have such a geyser. One thing I remember was that in the winter we got hot water. In the summer we got VERY hot water.

Another thing I remember was that at the end of one very hot day the temperature and thus the pressure in the geyser got too high. The geyser had a relief valve that popped open at a certain pressure - to protect against this condition. I had a geyser full of very hot water run down my roof and into the gutters.

Edited October 13, 2025Oct 13 by Bobster.
Sppeling

8 hours ago, Bobster. said:

I think the reference is to a certain type of geyser that heats water using heat from the sun and that heating does not use electricity in any way.

I used to have such a geyser. One thing I remember was that in the winter we got hot water. In the summer we got VERY hot water.

Another thing I remember was that at the end of one very hot day the temperature and thus the pressure in the geyser got too high. The geyser had a relief valve that popped open at a certain pressure - to protect against this condition. I had a geyser full of very hot water run down my roof and into the gutters.

I wasn't going to comment, but hell here goes. I've had this type of chat also with someone who owned a thermosiphon geyser in Graaff-Reinet, so no electricity involved, but the same principle applied. Leave the heating unchecked, and at some point the water boils to the point where the geyser relief valve opens and you dump a load of boiling water.

Now sure, you could have this type of split-system where you have a Kwikot geyser as per normal in the roof, and then a retrofit set of tubes, with the circulation pump powered by a small solar panel.

2 hours ago, tony Lampard said:

"I think the reference is to a certain type of geyser that heats water using heat from the sun and that heating does not use electricity in any way." Perhaps you can tell me how the Kwikot solar water heating retrofit pans out?

In this scenario you could turn the Eskom feed off, and if you've over-sized the panel relative to the geyser volume, you'd have the same problem in principle of excessive heating. The same issue potentially as putting the panels directly to the AC element without a controller.

Turning to the scenario here where 4x450W panels are hooked up to the 3kW geyser, consider that in summer with effectively 7 or more daylight hours you could generate up to 12kWh of electrical energy, which gets dumped straight into raising the water temperature On a 200l geyser that might be okay, but on a 150l geyser that's enough to raise the water temperature for an ambient 30 degrees C to boiling point. Even if one person hasn't had an issue in 3 years, I still wouldn't suggest it as a general solution for every household. Lots of folks could be off on holiday over Christmas, and having your geyser reach boiling point every day while you're out of town is just - well maybe nothing happens - but I still wouldn't do it.

For my money, back to the original title of this thread, being about a 5kW hybrid inverter, I see the Powerforum Store is now pitching the 5kW Solis S6 Hybrid inverter on special at R11,730 incl VAT. https://powerforum-store.co.za/collections/inverters/products/solis-s6-pro-5kw-advanced-hybrid-inverter

I'd start there with a 2kW element for heating by day. Around 3kW (6x550W) of panels for the 12-15kWh of daily consumption. And considering that the recent poster was asking about usage that's for the evening and also largely in the morning, I'd expand later with a heat pump for the morning usage, and 10kwh battery storage, and making the whole house more self-sufficient. All with CoC. To each his own.

42 minutes ago, tony Lampard said:

There seems to be some misunderstanding. The 4 x 450w panels go double break through a contactor (this switches both the + and - off in going through and back again through). The contactor is switched by a 220v AC coil and this control circuit is fed from the inverter through the thermostat. The song and dance has been about DC arcing and "how dangerous" that will be if the contacts weld closed. If that happens it just means the thermostat is no longer in circuit so the water could boil - thats a feature of the thermal solar water heaters anyway and thus the Kwikot "solarize" retrofit could have the same symptoms. 3 years down the line I have had no sign of arcing (so the geyser is always controlled by the thermostat).

Also the concept of "soft load" doesn't seem to be understood. If you have a 3kw geyser off a 5kw inverter that is nearly all of your max power taken up with one big smack - overload risk, even dedicated, is very high.... and this is poor optimization of the available energy. Direct from the panels the energy is optimized because the varying voltage changes the actual wattage (I gave you the equation) - sure the low voltage efficiencies aren't great but not totally wasted. This concept of "soft" load direct from the panels is also generating interest with DC motors, especially pumping water. The "soft" load is a variable speed drive speeding up and slowing down is regulated by the intensity of the sun; that takes the need for Demand Side Management (DSM) right off the table.

The fire risk is not with welded contacts that will keep on powering the geyser. It is the arc during breaking that will cause a flame that creates the heat and can melt the whole contactor. DC contractors are expensive for a good reason. It is not worth the risk of fire or claim rejected as it is not to the SANS code of wiring. Soft start or not.

Sadly, Tone's nonsense has hijacked this thread ☹️

Let's get back to the original discussion about advice regarding the installation of a smallish PV system on a tight budget...

I agree with the wise statements made by @Denns @Scorp007 @Bobster. and @TaliaB.

My suggestion would be to initially install a slightly larger than planned inverter and more than planned panels, and only add batteries at a later stage if required (but have the installer do the wiring, DC isolators, bussbar, etc. now. Reasoning for this is that the difference in cost between (say) a 5kW inverter and an 8 (or even 10) kW inverter is much less in terms of R/kW (a quick search show R20 000 for a Sunsynk 5kW, vs R30 000 for an 8kW of the same brand, vs R35 000 for a 10kW). PV panels are the cheapest part of a solar system, and here I would also opt for the slightly more expensive higher capacity units (550W - 600W), as here too the cost difference (R/kW) would be quite small.

PS. I'm not recommending the Sunsynk brand, only using their rather high prices as a comparison.

1 hour ago, HennieL said:

Sadly, Tone's nonsense has hijacked this thread ☹️

Let's get back to the original discussion about advice regarding the installation of a smallish PV system on a tight budget...

I agree with the wise statements made by @Denns @Scorp007 @Bobster. and @TaliaB.

My suggestion would be to initially install a slightly larger than planned inverter and more than planned panels, and only add batteries at a later stage if required (but have the installer do the wiring, DC isolators, bussbar, etc. now. Reasoning for this is that the difference in cost between (say) a 5kW inverter and an 8 (or even 10) kW inverter is much less in terms of R/kW (a quick search show R20 000 for a Sunsynk 5kW, vs R30 000 for an 8kW of the same brand, vs R35 000 for a 10kW). PV panels are the cheapest part of a solar system, and here I would also opt for the slightly more expensive higher capacity units (550W - 600W), as here too the cost difference (R/kW) would be quite small.

PS. I'm not recommending the Sunsynk brand, only using their rather high prices as a comparison.

The price difference between inverters is a lesson I learnt in December last year, unfortunately. I bought my 3.6kw inverter for 3.5k. Only to find out I could have gotten a 5kw for 5k or for 6k, a 6.2kw. If I upped the budget to 12k, I could have had an 11 to 12kW inverter. I had the money, but I had assumed that inverters run into the tens of thousands. I am basing pricing on the Voltronic clones, like the MUST and ECCO brands. Not the more expensive and more popular/reputable brands.

The inverter runs my whole home fine and hasn't had any issues with it in the year I have had it. I can pull close to 4kW from the panels, and it doesn't struggle to run anything in my house as long as I manage my loads.

Btw, I don't think the original poster expected his thread to be continued like this and have discussions about geyser heating. I am sure he has since grabbed popcorn and watched the debate between @tony Lampard and everyone else here🤣. I like the difference in opinion, though he is doing something unconventional, and it seems to be working for him. I might have done the same, but it's a risk I wouldn't be willing to take, primarily because of insurance and the slight potential danger to house occupants it may have.

Quoted

"The rationale is straightforward: direct solar-to-load consumption is 100% efficient, whereas battery charging involves round-trip losses (typically 10-15%). Why waste that energy converting and storing power you're actively using?"

Are you very sure that you can convert DC from panels to AC to the loads. I think most manufacturers that run at about 93-98% would like to have it at 100%.