PV Divert to immersion difficulties

[S2D2]

8 minutes ago, Radian said:

someone decides to bake a cake

This is optimal PV divert benefit

[Radian]

2 hours ago, dpmiller said:

I wouldn't overthink it tbh, they just... work.

That would be a refreshing change!

 

2 hours ago, S2D2 said:

This is optimal PV divert benefit

I'm doing my best to encourage such 'optimal usage' and am quite pleased with the response so far. Fresh hot cakes cooked by the Sun are a delight and reasonably guilt-free. But if I baked them it would still add to the carbon problem.

[ProDave]

We have had such a good run of sun this week my immersion maxed out mid afternoon.  I would (for once) have loved someone to take a long shower.

[dpmiller]

Same here, yesterday. Whole tank above 90c...

[S2D2]

£17.04 in export payments for the previous three days, £3.06 in gas hot water usage. Important to remember it's an unusual few days and not get tempted into investing in a divert system based on the "worst" case, it's the argument solar installers use to sell more unnecessary batteries too.

[Radian]

Terrific. Woke up to the RCD tripping off the house power just before 7AM. A quick look at my power plots confirmed my suspicion that the my Solar PV diverter had just started to dump the excess into the immersion. A megger test reveals that the insulation resistance of the immersion heater element was less that 1M. I can't seem to get a break with this project.

[Radian]

The faulty immersion has been replaced under warranty and, for the time being, extra expense on the plumbing has been vetoed while other more pressing issues are being addressed. This means I'm trying to make the most of what I've got: A 0.6m long immersion in the top half of a 1.2m tall (160l) cylinder. I've put in a permanent temperature sensor right at the top in addition to the one 1/3 the way up that was governing the boiler demand so I can plot the goings-on in the cylinder.

 

Now the problem remains being how to stop the boiler firing-up when the partial draw-off of lovely hot Solar PV diverted water is replaced by unmixed cold when there's every prospect of there being sufficient 'free energy' up at the top of the tank to make this unnecessary. Take today so far for example:

                      A nasty case of stratification: HW LO falls after 7AM while HW HI goes up.

 

Ignoring the shenanigans until the Sun comes up at 7AM, the immersion gradually builds up the water temperature in the top half of the tank (red: HW HI), even while the lower half loses heat through ambient losses and occasional draw-off (orange: HW LO).

Gas boiler demand is currently governed by the average of these two temperature readings (orange dashed: HW Average) so that at 12:30PM a 10 minute shower (🙄) displaces a significant amount of water causing the bottom sensor to fall to 20^oC thus pulling the average below the HW Setpoint.

The immersion was going great guns at this time and the HI sensor in the top of the tank shows plenty of HW was available - but still the boiler fired up.

 

I guess I could weight the two sensors differently - 50/50 seemed like a reasonable starting point. But taking extreme cases to aid understanding, if the sensing was biased to 100% of HW HI, there might only be a tiny bit of HW just 'floating' at the top and nearly 160l of room temperature water below it.

I could move the LO sensor up a bit but the current position has worked well before I built the PV diverter and in the winter its height reflects how long it takes to reheat the cylinder during sustained HW use. Any suggestions?

[ProDave]

Simplest way would be on a day you know it is going to be sunny all day, manually turn the boiler off.

 

Or have a summer boiler program, that turns the boiler off during the daytime.  This is what I am trialling at the moment,  we still need some heating (average about 5 degrees outside today!!!!) so I only have the heating timed to come on for a few hours around mid day when it is likely there might be some good PV generation to power most of the ASHP demand.

[S2D2]

56 minutes ago, Radian said:

The faulty immersion has been replaced under warranty and, for the time being, extra expense on the plumbing has been vetoed while other more pressing issues are being addressed. This means I'm trying to make the most of what I've got: A 0.6m long immersion in the top half of a 1.2m tall (160l) cylinder. I've put in a permanent temperature sensor right at the top in addition to the one 1/3 the way up that was governing the boiler demand so I can plot the goings-on in the cylinder.

 

Now the problem remains being how to stop the boiler firing-up when the partial draw-off of lovely hot Solar PV diverted water is replaced by unmixed cold when there's every prospect of there being sufficient 'free energy' up at the top of the tank to make this unnecessary. Take today so far for example:

                      A nasty case of stratification: HW LO falls after 7AM while HW HI goes up.

 

Ignoring the shenanigans until the Sun comes up at 7AM, the immersion gradually builds up the water temperature in the top half of the tank (red: HW HI), even while the lower half loses heat through ambient losses and occasional draw-off (orange: HW LO).

Gas boiler demand is currently governed by the average of these two temperature readings (orange dashed: HW Average) so that at 12:30PM a 10 minute shower (🙄) displaces a significant amount of water causing the bottom sensor to fall to 20^oC thus pulling the average below the HW Setpoint.

The immersion was going great guns at this time and the HI sensor in the top of the tank shows plenty of HW was available - but still the boiler fired up.

 

I guess I could weight the two sensors differently - 50/50 seemed like a reasonable starting point. But taking extreme cases to aid understanding, if the sensing was biased to 100% of HW HI, there might only be a tiny bit of HW just 'floating' at the top and nearly 160l of room temperature water below it.

I could move the LO sensor up a bit but the current position has worked well before I built the PV diverter and in the winter its height reflects how long it takes to reheat the cylinder during sustained HW use. Any suggestions?

Use either the current immersion power or the forecasted solar generation for the rest of the day as your blending coefficient. If you want to get really fancy model the current energy in the tank and the current loss rate.

[Radian]

3 hours ago, ProDave said:

Simplest way would be on a day you know it is going to be sunny all day, manually turn the boiler off.

 

Or have a summer boiler program, that turns the boiler off during the daytime.  This is what I am trialling at the moment,  we still need some heating (average about 5 degrees outside today!!!!) so I only have the heating timed to come on for a few hours around mid day when it is likely there might be some good PV generation to power most of the ASHP demand.

 

I can easily change the HW setpoint - I just bark at Alexa: "set hot water to zero" (or any other desired temperature). This has been one of the more useful hacks I've done with my heating system as anyone in the house can give it a boost when planning to use HW without opening an app or using a wall mounted control. The programmed values are then restored at the given times.

 

So yes, I can effectively override the boiler and even change the schedule but being lazy, and being an engineer, means I'm looking for the maximum amount of automation.

 

3 hours ago, S2D2 said:

Use either the current immersion power or the forecasted solar generation for the rest of the day as your blending coefficient.

 

This is a good suggestion. To begin with I will try it out by shifting the weighting towards HW HI as more PV divert Joules are accumulated.

 

4 hours ago, S2D2 said:

If you want to get really fancy model the current energy in the tank and the current loss rate.

 

I'll have a ponder on that.