Still we think you could do better. Insulate that geyser pronto!
I did a test last night on my 100l Kwikot B+. I have a Sonoff on the geyser - nothing fancy just resting against the flange. On the roof, no fancy lagging or blanket (no extra insulation at all actually ;))
When I stopped heating at 15:16 yesterday the Sonoff temp was 57.6C. It is now 46.1C and the minimum temp last night here was 4.9C.
So with some extra insulation my standing loss should drop a bit. I need to insulate the temp sensor as well!
So a 11 degree drop during a very cold night. In my layman’s mind that’s actually not bad and meaning that geyser’s are fairly well insulated.
But using all that fancy maths again (Phill please help) how much energy would it take too heat it back up to 57 degrees?
4.2kJ * 11.5 * 100 = 4830kJ
and 4830kJ / 3600J = 1.34Kwh?
Is this correct?
Unless I’m missing something from @Phil.g00 's maths there. Other than keeping the water hot for longer, the plan to add parrafin increases the standing losses. In the example it went up from 2kWh/day to 4kWh/day. Surely that means those losses need to be replenished every day, increasing your energy requirements even though it is more efficient. The extra efficiency comes from storing more energy in the paraffin and the losses become a lower % of the total energy stored…
Long reading 
But the few factors I took to heart.
The first is that for the normal geyser, there was a “cut-in” time of around 6 to 9 hours. So about two or three times a day, it turns back on to get the water back up to temp. If you put this into a PCM, then of course that doesn’t have to happen as often. But getting the temperature back up the next day, which is now going to be as much as 18 hours later, will of course take a bit more energy. That’s not too surprising.
The second is that because of the higher sustained temperature, your loss to ambient is going to be slightly higher. So you may keep the water a bit warmer, but the overall losses is going to be a little higher too. Of course that may not matter… it is still significantly more efficient than heating water with battery power. You’re using the sun.
I’ll read more and reply again later
Maybe, you’d think so.
It depends on the position of the thermostat. Which is what I am trying to get my head around.
This is what I am trying to get my head around. The thermostat should read the temperature of the water. You do realize that keeping the thermostat warm is not the same thing as keeping 100l of water warm.
Either the thermostat is accurately indicative of the water temperature, or it isn’t.
Hold it over a candle. or in your case, it sounds like put it in direct sunlight or shade, it won’t make a blind bit of difference to the actual water temperature.
Claiming that your water temp is X, and in the next breath say the thermostat needs to be fitted correctly is an inconsistency.
Can’t help with that Maths sadly 
I’m just happy I got to 57 with my 2kw element on a cloudy day from my PV alone - no grid involved. We use gas in the mornings and PV heated water in the evening so works for us.
That’s impressive! (I presume this is a 220V system?)
How do you manage your PV power to only power the geyser in the afternoon?
And what about other loads
I have a Sonoff driving a relay for the geyser using Node Red. Only switches when SOC greater than 96%.
Yes: 220V from my MPII 5kw.
Naaa, the differences, if any, would be cents on the rand, the rand coming from a 2kw element on during peak solar hours.
This is true, overall in the PCM model, there are more absolute losses, although the overall process is more efficient.
I actually assume the losses are double. (I don’t know if that accurate or not).
The distinction is I get to add this extra energy from direct solar, as opposed to adding it from my batteries.
This is a thermal battery. Just like a small electrical battery when it’s full your solar shuts down, the excess power goes to waste. This is just a bigger battery, and like a lithium battery it has a flat discharge curve that keeps its voltage ( read temperature) stable for a long time as it discharges.
Again true, but I do make that distinction. (Reread Note 1). The hot water in the referral paper you mentioned wasn’t actually the same thing as heating water and then leaving it to cool.
(It wasn’t left alone, but maintained). I am aware that cooling isn’t a linear plot.
If I recall your referral paper, caters for losses where the thermostat kicks in and maintains the high temp, just like in my PCM geyser. So in that sense, the loss rate is going to be very similar in both geysers as they are both at the higher temp.
On the other hand, if you were to use the extra energy to overheat the water in a standard geyser. This in an attempt to have hot water for longer. The losses would be far greater. and the process even more inefficient for the reason you state.
This is also what puts a song in my heart. Our geyser works only twice in a 24 hour period. One of these “recharge” sessions is done completely by PV and no grid. Big win for us.
But… I doubt whether I’ll be able to manage this on all cloudy days. It depends on the type of cloud cover I guess. Sometimes I manage this, other times not.
Some of us have waaay less loads to manage so its quite easy. 
Ours also works, evenings and early mornings, and that from one heating session per day.
I understand. Effectively you give yourself a “heat” battery instead of an electric/chemical one. I guess what would be more efficient is if you “charge” you “heat” battery directly from solar, and not from an electrical element.
So maybe use a solar collector rigged for paraffin to heat your “heat” battery around the geyser. Like an EV-tube thing? I like the principle of converting energy less if possible. Electricity is nice because we have so many devices that can convert it into other forms, but ultimately there would be losses.
Also, then you wouldn’t need to care about where to put your thermal probe. 
Yes, this is true. I assume the absolute losses have doubled. The distinction is these energy losses are catered for by sunshine during the day, not batteries during the night.
In my layman’s mind I completely understand this. Which is why I know my “tests/observations” is very rudimentary and probably inconsistent at times as well.
But a temp probe that sits on the very same spot, day in and day out every day giving a reading of say 50 degrees whilst the actual water temp is 60 degrees, should give those readings every day all other factors remaining fairly constant?
Then who cares about temp sensor placement and the 10 degree difference?
All I care about is that I’m physically aware of reduced heating time, and then even this energy saving could be calculated despite the temp probe and actual water difference using either of those values.
I know there are many more factors at play, but this is easy enough for me and my application.
Even better! I’d like to get to this point as well. But it would entail a second geyser. Alternatively unless we only shower and do that military style.
Well, I don’t know if this is true now.
If it is reading consistently 10 degrees low, then yes what to say is correct, you are comparing relative temperatures.
But, if this temp robe is merely attached as you volunteered, it may be subject to direct sunlight in the am and shade in the pm. (or vice versa). If it is in ceiling, then the ambient will be extremely hot in the day and extremely cold at night.
I instances like that, it isn’t consistently a relative indicator of the true situation.
Hence, I question the validity of your conclusions based on your own account.
Yes true, I get this.
In my specific circumstances though, I don’t have an additional temp sensor, and my geyser never ever sees sunlight or gets rain on it.
It just lies there snug in its little brick house blissfully unaware of its surroundings.
And where did you question my conclusions hahahaha. Don’t think you have.
But thanks a lot, I value your significant input here, certainly am learning a lot!