Earth Notes: Octopus Heat-pump Journey

It is ~6°C outside and is due to be similar over the next few days. I have turned the boiler flow temperature down to minimum (~35°C) to see if the house is warm enough, for a weather compensation data point. We barely need heat at ~12°C externally, so that is another data point.

The boiler seems to be running in bursts of 1 minute or shorter, maybe ~20% duty cycle, with two radiators ~fully open (out of six) and most of the rest slightly open.

As of ~21:30Z and nominal end of the heating period, I think it has worked well.

2024-01-13: up a bit

TRVs 1g and 3l were not turning off, though the rooms were not cold, so the ~35°C flow temperature may have been marginal for those rooms at the ~5°C outside at ~8pm on the 13th.

Temperatures are due to drop significantly over the next few days (to below freezing overnight), so I have put the flow temperature back up to a more typical ~45°C. At this level 1g and 3l are usually turning off before the end of each heating period on 2024-01-14.

2024-01-16: winter design temperature

The weather forecast claims -3°C outside, my porch sensor says below -1°C. Thus near enough London winter design temperature of -2°C.

Although slightly tricky to tell given all the comings and goings before the heating goes off at ~9am (on at ~07:30), the 45°C flow temperature seems to be sufficient: rooms are coming up to temperature and TRVs are turning down.

2024-01-18: winter design temperature or lower

The weather forecast claims -4°C outside, my porch sensor says near -2°C. Thus near enough London winter design temperature of -2°C again. (The hallway Radbot on frost protection ~6°C came on this morning...)

The living room is struggling a little to get up to temperature (with the RadFan off), so I have opened up the lockshield a little.

2024-01-19: winter design temperature or lower, again

Close enough to London winter outside design temperature (-2°C) again. Towards the end of the 07:30 to 08:59 nominal morning heating on-time, only the living room (4 on Radbot, ie ~21°C target; RadFan off) is still calling for heat.

(Mains water temperature was ~9°C.)

The Octopus survey is due soon. A heat-loss calculation will need to be done, during the survey or in any case before an installation. To that end, I am gathering data to help with that process and calculation.

• 1960's timber-framed end-of-terrace ~76m^2 over two floors, 2024-02-09 EPC A 100.
• ~300mm loft insulation (Octopus later rated this as U=0.12/W/m^2K).
• Smallest bedroom (and other bedrooms) 30mm aerogel external wall IWI, U=0.3W/m^2K.
• Front and back (Ug=1.0W/Km^2 with SolaGlas Climacontrol Vertical) doors both new-ish good double-glazed.
• All windows triple glazed Uw=0.8W/m^2K.
• Living room 40mm aerogel external wall IWI U=0.21W/m^2K, and insulation above the ceiling to 0.33W/m^2K.
• Kitchen 40mm aerogel external wall IWI, U=0.24W/m^2K.
• Downstairs solid floors, U=0.63W/m^2K.

The Octopus survey has landed; their tool's suggested heat loss (50°C flow, and presumably at ~-2°C design temperature) is under 4kW (3805W) for 16WW, though their estimated heat loss for the living room of a little under 1kW feels a bit high to me. But I don't mind upgrading the living room radiator, since it clearly is a bit marginal; a K2 vertical 1800mm might do the job. Their calculations also suggest that the hallway radiator is ~20% undersized given that this is nominally providing heat for the upstairs landing also, but it is barely ever even on, so I am contesting that! Their calculations also suggest that the kitchen radiator is ~20% undersized, but a 600mm high rad in the same spot and the same length would pass muster. Again, in practice, that room is never cold, likely in part because of incidental gains from the gas boiler, Thermino, cooking, and dishwasher. At one point the solar PV and Enphase battery had to be turned off for a few minutes so that a load test could be done.

I asked for maximum data access, and possibly control, to better make the system grid-responsive and low carbon. I asked for heat flow metering of space heat and DHW (so I can do H1, H2, H3, H4 boundaries)! I asked for a way to do the data without it leaving the house (like the Enphase, unlike the Eddi).

The survey took two people (including a trainee) approximately 4 hours.

It seems that everything we need can fit, physically. I will need to do some work to see how the end result can be made to look good, and allow for easy maintenance. One initial idea is to effectively rotate a cupboard and a half to match the existing ones on the same wall, on the other side of the fridge/freezer.

Looking at the spec for the suggested DHW tank, Joule Cyclone Slim High Gain Solar 200l - TCIMVG-0200SFC, it seems to have an actual capacity of 188l (similar to the Thermino) but a heat loss of ~2kWh/d, so about twice that of the Thermino.

Given an expected DHW CoP of ~2 for the heat pump, and our typical ~4kWh/d DHW demand, it seems that resistance heating (eg in the Thermino) would have similar energy footprint to using the heat-pump, because through the raw CoP is ~2 we lose half the heat pumped thus back to ~1.

In summer that may be true, ie let the Eddi divert excess PV. (Maybe first to the cylinder then the Thermino to provide a ~2d store.)

In summer that may be true, ie let the Eddi divert excess PV.

But in the winter, lost heat from the DHW would be warming the kitchen, and in particular if we can avoid heating the full tank if not needed and maybe heat only to ~50°C, so as to reduce losses, then the heat pump set up should be providing some gain over straight resistance heating, reducing grid demand. It still feels like a lot of space for a relatively marginal gain, given that the Thermino is already there!

2024-07-07: more thoughts

Assume a simple consistent 4kWh/d of DHW demand, a Thermino resistive CoP of ~1 and a heat-pump DHW CoP of ~2. Assume losses from the Thermino ~1kWh/d, from the proposed water tank ~2kWh/d.

Initially ignore where the electricity comes from (PV diversion or grid), as any we do not use because of better efficiency can spill to grid to make the grid greener.

Assume for now that tanks are generally kept full-ish, though less so during grid intensity and demand peaks. Either the tank or the Thermino can cover a whole day's demand, so when in series the (upstream) Thermino usually only need cover its own losses (assuming that the Thermino is only heated when once the water tank is full and a big lull in available green energy is predicted).

I am thinking about three ways of setting up DHW:

• Hot water tank only: use + loss = 6kWh/d, so 3kWh/d electricity demand.
• Thermino only: use + loss = 5kWh/d, so 5kWh/d electricity demand.
• Thermino as pre-heat for hot water tank: 7kWh/d, so 4kWh/d electricity demand. (This provides up to two or three full days' supply, so allowing better times to top-up to minimise footprint.)

If we had the heat-pump only provide space heating, and so saved the space for and upfront carbon of the water tank, we would use ~2kWh/d more electricity (thus an extra ~700kWh/y). That extra will be increasingly low carbon as the grid is. In practice most would come from our PV so maybe we would import and pay for (and not export, and thus not be paid for) ~300kWh/y costing maybe £100/y, for the remaining ~5 year life of 16WW. So if omitting the tank saves over ~£500 we should probably do it.

Tuning the heat pump to maximise space heating efficiency for its ~2MWh/y demand should be the primary goal.

2024-07-15: moar thoughts

Losing 2kWh/d from heat-pump heated DHW in the main tank at CoP of 2 is pretty much the same as losing 1kWh/d from the Thermino at a CoP of 1, in terms of carbon emissions.

So I am leaning towards keeping the tank and the Thermino, embracing the extra ~36h storage, and topping up the tank/Thermino resistively from grid only when super-green (ie bottom quartile by intensity over last 7d, and storage not being emptied), or possibly when intensity is below mean and a Dunkelflaute is forecast.

Octopus called this morning and seems prepared to go ahead. It sounds as if there are enough installers locally that installation could happen fairly quickly, eg within ~6 weeks, if required.

I have asked for some more detailed information on equipment so I can be happy that everything would fit.

I have asked for H3/H4 boundary data and heat metering, preferably compatible with OpenEnergyMonitor systems.

An earlier email said that Octopus would not fit heat metering.

I have also confirmed that we don't offer H3 or H4 boundary data and heat metering. However, you can buy that yourself and fit it at a later date as this isn't a service we provide currently.

Action (metering): have Octopus confirm that its installers will leave cabling and pipework easy to add electricity and heat metering.

Today I received a call telling me that design files were on their way for me to approve (assuming that my partner is happy to go ahead).

Apparently two heat-pump external units would work for 16WW: the Daikin EDLA04E2V3 suggested on the install design pack, and the Cosy 6 which has low-GWP propane refrigerant (and can run at up to ~70°C). I have also asked which has lower continuous power output (eg better modulation depth) to be able to reduce cycling away from mid-winter.

I will double-check that everything is able to fit in a 1m x 66cm footprint in the corner of the kitchen where the combi currently is. In the latest Octopus email I am told that:

Please note that the size of the buffer is about 1m tall by 300mm diameter, the expansion vessel is about 400mm tall by 300mm diameter, the diverter valve is small about 150mmx1500mm.

As of late August I have my other half's agreement to go ahead providing I can make things look OK in the kitchen, and no decor gets wrecked elsewhere.

Before I left for a short break with my family, I sent Octopus a longish list of questions. I will need to be happy with these points before I sign off an Octopus design, on the back of which Octopus applies for the BUS grant.

1. Do you (Octopus) supply and fit a TMV (Thermostatic Mixing Valve) on the hot water output to the rest of the house to limit DHW to ~50C?
2. Can you please confirm that your installation team will configure the input to the cylinder to be plumbed from the output of my Thermino heat battery, ie to use the Thermino as pre-heat for the DHW cylinder.
3. Can you please confirm that the Thermino 150 + cylinder + buffer tank + all other works will fit within the 100cm x 66cm footprint available, so that I can provide an acceptable final enclosure.
4. Can you please confirm that electricity consumption of the heat pump is either available in real-time eg locally or over the Internet, and/or that your installation team will configure the system so that I can easily fit accurate electricity consumption monitoring myself (within the 100x66cm footprint)?
5. Can you please confirm that your installation team will configure the system so that I can easily fit heat metering for the DHW and space heat flows (within the 100x66cm footprint)?
6. The cylinder has a space for an immersion heater.
   1. Will the immersion heater be fitted?
   2. If (6a) is yes, at what rating (kW)?
   3. If (6a) is yes, does the heat pump control the immersion heater to run a legionella cycle, or can it be wired to my Myenergi eddi so that I can use it for solar diversion and a regular legionella cycle?
7. Is temperature measured at multiple points in the DHW tank?
   1. Will I have access to those temperature readings and/or a percentage-full metric in real-time?
8. Would you please note that for reasons discussed with your survey team we would only be taking the living room radiator upgrade (for example the hallway radiator is only turned on about 24h per year).

I will probably also need more detailed instructions for the trench works for my builder.

As (8) I may want to incorporate some of the answers to the above into your formal offer to me, or note them is a side-letter as additional conditions of sale.

To clarify one thing: when I am asking for access to data I do not mean via an app (I do not run the Google App store for security reasons), I am asking about an API that I can use from the Raspberry Pis that run my systems, eg as I use to manage the Myenergi Eddi.

Octopus answers, reformatted as above:

1. We don't supply and fit a TMV (Thermostatic Mixing Valve) on the hot water output to the rest of the house as the water is heated to and stored at 50 degrees
2. No, we won't be doing that, as Thermino isn't compatible with our heating system.
3. Being that we will not be installing the Thermino, I honestly cannot confirm if it will fit within the 100cm x 66cm footprint available
4. I spoke with my design team and they confirmed to me that the Daikin App gives all the information needed. Coincidentally, we do not offer API support. It was however suggested that you can install a CT clamp to measure the usage.
5. We cannot stop you from integrating heat metering. However, I was informed that If you are going with a proper sensor, it'll involve tampering with our installation which will void your system of warranty
6. The immersion heater will be fitted and is 3kW. Once every week, the immersion heater raises the temperature of the water in the water tank to 65c and flushes itself to ensure that you are not exposed to legionella disease.
7. No, the thermistor in the cylinder is a third way or almost in the middle; this is where the temperature will be measured.
   1. All information are available on the Daikin App. We don't offer API support. I have attached the brochure for the Daikin system to provide you with more information about the Daikin system.
8. Our installation needs to meet MCS standards. Once the radiators have been fitted, you can turn them off using the TRV. However, if you want to keep your radiators, you would have to fit an electric radiator to ensure 97% of the heat loss in the rooms is covered. Please note that Octopus Energy will not bother you with how you run the heating but must ensure for audit purpose that we meet 97% of the heat loss of your property.

I work through these, and Octopus answers, and my responses, below.

Octopus has also confirmed by email that I can confirm based on the heat loss of your property being 3805 W, you qualify for the cosy 6 heat pump.) My other half is OK with the look of the Cosy 6, and I like the low GWP (Global Warming Potential) of its refrigerant. MCS product directory: Octopus Cosy 6 SCOP 3.98 at 35°C, 3.21 at 50°C, 2.72 at 65°C. Other ASHPs get to a little over 4 (eg 4.5) at 35°C and over 3.5 at 50°C.

Action (Cosy): have Octopus confirm that a Cosy 6 unit will be available if required CompletedActionStatus.

Thermostatic Mixing Valve (TMV)

(1) Do you (Octopus) supply and fit a TMV (Thermostatic Mixing Valve) on the hot water output to the rest of the house to limit DHW to ~50°C?

Octopus answered:

We don't supply and fit a TMV (Thermostatic Mixing Valve) on the hot water output to the rest of the house as the water is heated to and stored at 50 degrees.

This seems to me potentially at variance with building regs, water regs, and HSE advice, as during legionella cycles (heating to 60°C) water may be distributed hot enough to scald, which our current system avoids.

This may determine how the Thermino and the heat pump have to work together, see the next question/answer.

Action (scalding risk): I have asked Octopus to respond on this potential scalding issue (now intend to use Intasol to using tank as if was the combi to avoid this risk CompletedActionStatus).

Interaction with heat battery

(2) Can you please confirm that your installation team will configure the input to the cylinder to be plumbed from the output of my Thermino heat battery, ie to use the Thermino as pre-heat for the DHW cylinder.

Octopus answered:

No, we won't be doing that, as Thermino isn't compatible with our heating system.

This is a problem because at the very start of the process I was told that an Octopus heat pump system could work with existing solar DHW including in the form that mine is (solar PV diversion into a heat battery). I am certainly not keen to scrap my Thermino installation to accommodate a heat pump: I see the extra energy storage (and grid decoupling) as a definite plus.

There are two obvious ways to configure the Thermino and heat pump to work together:

1. as above using the Thermino as pre-heat, plumbing the units serially, with the water feed as cold mains -> Thermino -> DHW tank -> house
2. else connecting the heat-pump tank the same as the combi currently is:
   • if the Thermino is hot enough the Intasol takes DHW from the Thermino
   • else if the Thermino is too cool, the Intasol routes to the house from the DHW tank and bleeds remaining Thermino heat in to the DHW tank (down-mixing flow to the tank to 28°C may or may not be beneficial here).

In each case the Thermino is topped up by diversion and/or when grid carbon intensity is especially low.

If Octopus declines to fit a TMV at the output of the DHW tank then the first option is not safe, in part because the Thermino output can reach ~80°C and could be routed through the DHW tank at that temperature, but even the heat-pump legionella cycles will hit 60°C.

Rearranging the current system to substitute the DHW tank for the combi as in the second option protects against scalding whatever the source of very hot water.

Action (solar heat battery de/re-commissioning): have plumber/etc disconnect and move Thermino and Intasol before Octopus installers arrive, then reconnect afterwards, in second configuration.

Action (solar heat battery arrangement): have Octopus confirm that its installers will leave pipework easy to reconnect the Thermino and Intasol per the second configuration option.

Will everything fit in the available space?

(3) Can you please confirm that the Thermino 150 + cylinder + buffer tank + all other works will fit within the 100cm x 66cm footprint available, so that I can provide an acceptable final enclosure.

Being that we will not be installing the Thermino, I honestly cannot confirm if it will fit within the 100cm x 66cm footprint available

This one requires the equivalent of some careful sketching on graph paper to resolve enough to de-risk the install. I use LaTeX picture, files in a "2D sketch" working directory. I may move on to OpenSCAD for 3D.

(This is not the sort of effort that non-technical and non-determined users should have to make, and fitting into tight spaces co-existing with other uses is going to be a recurring issue with UK domestic retrofits.)

Arrangements:

1. Expansion tank mounted inside left side of cupboard: main disadvantage is need for cupboard to carry significant load; plan, side; working to 100cm x 66cm x 220cm
2. Expansion tank mounted above DHW tank: plan, side; working to 100cm x 66cm x 220cm
3. PREFERRED Expansion tank mounted above DHW tank: plan, side; working to 98cm x 74cm x 230cm

Selected inventory and spatial arrangement follows...

Action (tanks): have Octopus ensure that buffer and expansion tanks brought to site are no larger than the indicated 300mm diameter, and that the DHW tank is slimline and no larger than 475mm diameter (smaller would make life easier), and that placement of all the blue Octopus-fitted items as shown in the "arrangement 3" plan and side views is acceptable, not just the DHW tank.

Electricity metering

(4) Can you please confirm that electricity consumption of the heat pump is either available in real-time eg locally or over the Internet, and/or that your installation team will configure the system so that I can easily fit accurate electricity consumption monitoring myself (within the 100x66cm footprint)?

I spoke with my design team and they confirmed to me that the Daikin App gives all the information needed. Coincidentally, we do not offer API support. It was however suggested that you can install a CT clamp to measure the usage.

For security and other reasons I will not be running any mobile phone app as a critical part of a heat-pump system.

This triggers more questions to clarify:

• Can either heat pump be controlled meaningfully without a smartphone?
• Is the the case that a CT clamp would be needed to collect any consumption data for the Cosy but not for the Daikin?
• Does the Daikin 'app' run in a Web browser or only on a smartphone?
• Can the Daikin app data be accessed via a Web browser or automatically via an API?
• Can such data be accessed locally, or does it all have to go via Daikin servers?
• If Daikin's servers go down, or Daikin UK goes bust, does the heat pump fail?

Probably all I want on day 1 is reasonable control without a smartphone, and reasonable gross electricity consumption data. Everything else can wait.

Action (app): ask Octopus for more clarity on the Daikin 'app': in particular can either heat pump be controlled without a smartphone, and what happens if (say) Daikin goes bust or my Internet connection is down?

Action (basic consumption data): ask Octopus if any basic consumption data available from either heat pump without installing a CT clamp CompletedActionStatus?

Heat metering

(5) Can you please confirm that your installation team will configure the system so that I can easily fit heat metering for the DHW and space heat flows (within the 100x66cm footprint)?

We cannot stop you from integrating heat metering. However, I was informed that If you are going with a proper sensor, it'll involve tampering with our installation which will void your system of warranty

This is troubling, but there is already the metering action above, asking installers to leave cabling and pipework easy to add electricity and heat metering.

Action (warranty): ask Octopus for more clarity on the warranty issue: why would fitting heat metering violate any warranty compared with (say) replacing a TRV or fitting a decorative higher-output radiator upstairs, or reconnecting the DHW heat battery, for example?

Action (warranty DHW): ask Octopus to explicitly confirm on the warranty DHW issue: that recommissioning the solar/heat-battery part of our system will not not violate warranty.

Immersion heater

(6) The cylinder has a space for an immersion heater. (a) Will the immersion heater be fitted? (b) If (6a) is yes, at what rating (kW)? (c) If (6a) is yes, does the heat pump control the immersion heater to run a legionella cycle, or can it be wired to my Myenergi eddi so that I can use it for solar diversion and a regular legionella cycle?

The immersion heater will be fitted and is 3kW. Once every week, the immersion heater raises the temperature of the water in the water tank to 65c and flushes itself to ensure that you are not exposed to legionella disease.

Again, a number of new questions arise!

• What controls and drives the immersion? A signal from the external unit or something else?
• Can the timing of the pasteurisation run be adjusted, eg to only be in a ~2am to 4am slot when grid demand is usually low?
• Can the immersion instead be driven my spare Myenergy Eddi channel, so that I can optimise timing, eg avoiding high-carbon intensity times?

Action (immersion control): ask Octopus for more clarity on immersion control: in particular can it be controlled in a grid-friendly low-carbon way such as from my spare Eddi channel?

Temperature measuring

(7) Is temperature measured at multiple points in the DHW tank? (a) Will I have access to those temperature readings and/or a percentage-full metric in real-time?

No, the thermistor in the cylinder is a third way or almost in the middle; this is where the temperature will be measured. (a) All information are available on the Daikin App. We don't offer API support. I have attached the brochure for the Daikin system to provide you with more information about the Daikin system.

The exact tank model quoted (TRSMVH-0180SFC) in Octopus' proposal is not shown in the manufacturer's images and diagrams that I can find.

Action (sensors): ask Octopus if it can be made easy to place one or more additional temperature sensors on the DHW tank CompletedActionStatus.

All radiators?

(8) Would you please note that for reasons discussed with your survey team we would only be taking the living room radiator upgrade (for example the hallway radiator is only turned on about 24h per year).

Our installation needs to meet MCS standards. Once the radiators have been fitted, you can turn them off using the TRV. However, if you want to keep your radiators, you would have to fit an electric radiator to ensure 97% of the heat loss in the rooms is covered. Please note that Octopus Energy will not bother you with how you run the heating but must ensure for audit purpose that we meet 97% of the heat loss of your property.

Although it is a bit of a waste (especially for the hallway) this is probably not the hill to die on, and I think that we just take all the offered radiators.

Action (radiators): me to double check with partner that radiator upgrades are each acceptable (CompletedActionStatus): 2024-09-26 recalculation means only living room radiator upgrade will be needed.

Action (decor): stress to Octopus need to be especially careful to avoid decor damage on radiator refits.

Other actions

Action (power socket): maybe have electrician move/adjust the existing power socket to continue to be able to power the fridge, RPi, etc.

Action (colour): ask Octopus what colour finishes are available for the Cosy 6 as blue is not acceptable CompletedActionStatus.

In the Fediverse I asked:

@bazcurtis First questions if I may, while I plod on with trying to understand?

1) Do you use the Daikin 'app'? Does it run only on a smartphone or can it run in a browser?

2) Can you get any useful data out of the app such as basic energy consumption and CoP?

3) Can you get any of that data in an automated way?

4) Can you control the heat pump in any meanful way without a smartphone?

5) I'm going to [guess] that you went with OEM [OpenEnergyMonitor] because you couldn't get enough or good enough data without it?

To which he generously responded:

1 - I don't use the app much. I have all the stats in Home Assistant via ESPAltherma. Not sure about the browser

2 - Yes, you can get stats from the app, day, week year. I am not sure how accurate it is.

3 - I [don't] think you can get data automatically, but they do have an API. I also have that in Home Assistant. Look for Speak To The Geek on YouTube.

4 - The controller has a lot of settings.

5 - I like data and using ESPAltherma meant I could get more meaningful data.

Assuming for the moment that the heat pump goes in, and that the Thermino connects to the DHW tank as it did to the combi, then the control algorithm should change as the Thermino CoP of 1 means that the heat-pump should be used in preference whenever possible.

(The grid-frequency response mechanism is separate and can be left as-is.)

The Thermino should continue to nimbly accept solar diversion as now, though when there is sufficient solar generation over other house loads (maybe allowing for 1kW of Enphase capacity when available), then the heat pump should be forced to try to top up DHW for the better CoP.

The top-up (boost) driver algorithm could be simplified in the first instance:

1. locked out for a broad peak between about 16:00 and 22:00 local time,
2. locked out except when the grid is "supergreen" (in the bottom quartile of carbon intensity, with no grid storage being discharged),
3. maybe locked out if absolute carbon intensity is higher than from burning gas,
4. then allowed when current intensity is below ~0.5 of the recent actual 7d mean or the forecast mean for at least 2d.

For example, a paired (matched computation method) current value and a 2d forward average for the GB grid can be derived from the results of something of the form:

curl -s -X GET https://api.carbonintensity.org.uk/intensity \
    -H 'Accept: application/json' | \
    jq -c '.data[].intensity.forecast | floor'
147
% curl -s -X GET https://api.carbonintensity.org.uk/intensity/2024-09-03T13:41Z/fw48h \
    -H 'Accept: application/json' | \
    jq -c '[ .data[].intensity.forecast ] | add / length | floor'
161

Gentle Thermino ramp up and down, and possibly some proportional fill-target response to relative carbon intensity or pricing within the permitted regions, will be useful to retain/refit.

Also, a fallback setting in case of a problem with the heat-pump to keep the Thermino full outside peak time (else partially full) could usefully be retained.

No immersion control

All this applies even when on the assumption that the immersion heater in DHW is not controllable by me. In this case the Thermino input remains as now, driven from one Eddi output. The only topping up is to the Thermino.

With immersion control

The second control variant is for if I can control the DHW tank immersion from an Eddi output.

In this case I would make the DHW immersion top priority for diversion/boost, with the Thermino being the Eddi auto-secondary/fallback for diversion/boost.

I would change the legionella pasteurisation cycle to be 30 days after the last time the tank reached 60°C, whether by pasteurisation or diversion/boost. Much of the time no explicit extra pasteurisation would need to be run at all, and when drawing from the grid it would only generally be at low intensity. The pasteurisation would only apply to the the DHW tank.

This morning I received an email from Octopus saying that a complaint had been raised. Not by me!

It seems likely that someone in Octopus' social media team saw me posting in the Fediverse about the effort that I was having to put into the 3D jigsaw puzzle, checking that everything would fit in limited space, and raised a complaint internally. It was a bit weird...

It is indeed likely that many heat-pump receipients in the UK, especially those currently with gas combis and in small homes, are going to need a lot of help working how to slot everything into the tight spaces available. In my case the extra awkward element is getting my Thermino alongside the DHW tank, buffer tank and expansion tank.

I asked for the complaint to be closed, but to convey up the food chain the point that this 3D-juggling help likely being critical to get installs at scale!

Peer review

Our regular sensible builder (DM) came round today and we reckon that the internal space I have to work in is ~98cm (losing for the side panels, but gaining from no aerogel behind the combi) by 74cm depth by 230cm to the ceiling.

We will move the current double socket up above the Thermino for the RPi etc, and provide another double socket behind the fridge for the fridge and trailing socket.

An outline four-phase work plan looks like:

1. Dave M in to:
   1. Strip out all the existing cupboard, boxing (floor and ceiling) to boundary of new enclosure at least, move socket up behind Thermino and add new double socket behind fridge, move existing wiring as required, and set up anything needed so that the Octopus folks don't encroach on space we need for the cupboard and Thermino etc.
   2. Thread new T+E cable through for eventual DHW tank immersion control from Eddi.
   3. Do trenching, lay concrete base, and any other prep works needed for externals.
   4. Disconnect the Thermino / Intasol / expansion tank and make Thermino electrics safe.
2. Octopus does its thing: remove combi, install heat pump, DHW tank etc, radiator replacements.
3. David S (Beautifully Green) recommissions the Thermino / Intasol / expansion tank.
4. Dave M comes back and fits cupboard and makes good everywhere.

I may have scored an ancient OEM (Emoncms) system with 100A current clamps from a friend. So I could monitor total heat-pump electricity draw from the start, though I might add or improve monitoring later. That would be a more-or-less direct substitute for monitoring gas for combined space heat and DHW as of now. Later, separating those two uses may be interesting.

2024-09-09: Eddi

I have just woken up out of a dream to realise that I could use the spare current clamp slot in the Eddi with a new CT clamp (£15) or a Harvi to measure the heat-pump draw. It would not be fantastically accurate, but I do already have the data collection mechanism, and the Eddi would then the the conduit for all our heating energy. And it would avoid having yet another box somewhere...

Octopus has confirmed that it (the design team) is OK with the placements shown in the side and plan views (alternative 3), so I have signed thee provisional sales "Initial Design Pack" DocuSign form. Lots more details to be sorted out...

I have been handed on to the next person in the process at Octopus, who is now on holiday for a few days.

Their email includes:

...

Your ENA application - Octopus
We need to notify your electricity Distribution Network Operator (DNO) that you're having a heat pump installed. They will then inform us whether any work is required on your meter or on the local grid. To do this, I'll send them an ENA application on your behalf.

As soon as we've heard back from your DNO, I will be in touch to discuss the next steps (they aim to respond within 7-14 days). Alternatively, it's possible that your DNO will be in touch with you, so please keep an eye out for contact from them. If you do have any correspondence with the DNO please let me know.

Claiming the BUS grant - Octopus
I will be applying for the Boiler Upgrade Scheme (BUS) voucher of £7,500 for your heat pump install around 3 months before your install, this is in order to avoid the voucher expiring. Once the application is sent off, Ofgem will contact you in order to gain your consent for us to apply on your behalf. Please do keep an eye out for this, as it could end up in your junk mail.

HIES Workmanship Warranty
Finally, I will shortly be applying for your HIES Workmanship warranty. This is purchased to protect consumers from companies that may cease trading during the warranty period of your product. This policy will be sent to you directly from the insurer once your install has taken place.

...

Though I would like to directly control the heat-pump for grid interaction, I assume that at least initially that will not be possible, and it will have to run fairly autonomously. I may have to buy and integrate some extra hardware for such control..

Late last year (2023-11-24) central heating on times were set to be nominally 07:30 to 08:59, 13:00 to 15:59 and 19:00 to 21:29 (7h total). GB peak space/total heat demand is at ~6pm and peak ramp rate is at ~7am [watson2019peak], so both are avoided in this scheme.

I would still like the heating to come on with a call for heat from Radbots, and only to be enabled for some of the time as now. Maybe with a set back temperature (how measured?) rather than completely off outside those times, in an 'eco' quiet/efficient mode. Maybe the 07:30 to 15:59 period can be merged into one for simplicity. Maybe the period just before 4pm can have a raised setpoint per [berry2023flexibility] to let the house coast on thermal capacity a little. This is presumably not how it would be set up by Octopus.

(We do not want the heat pump running when we are trying to sleep, as well as avoiding grid peak times.)

Likewise, for DHW, I may be able to set a temperature profile against time, in which case it may be lower outside the heating-on times, and maybe raised for a little before 4pm to store a little extra energy. This should ensure that even taking a bath at peak grid time will not result in running out of hot water but should be more grid friendly.

Based on guidance tank temperature might be ~45°C when set back (eg during peak grid demand), ~50°C normally, ~55°C or whatever the heat-pump can reasonably achieve during boost, and 60°C+ for pasteurisation and to capture diverted solar energy, relying on the Intasol to limit temperature to the house to ~50°C; see also [levesque2004water].

I would also like to reduce legionella cycles to monthly, in the small hours, and preferably ~30 days since the tank last reached 60°C+, since it may be possible to top up from diversion or low-carbon grid juice opportunistically and thus avoid most explicit pasteurisation for its own sake; again see [levesque2004water].

Longer

Today I received an email from our DNO, UK Power Networks, approving our installation of a heat-pump.

It notes 16WW as being single-phase, as having a 60A main fuse, TN-C-S (PME) earthing, and not looped.

The heat-pump maximum demand is given as 36A, which would be ~9kW, presumably allowing for start-up transients, given that the device is listed as 4.35kW import.

Octopus did not receive (or did not use, in any case) my supplied building fabric information. Octopus used the following (generally worse) values:

• Walls - Brick with Timber frame - U-Value 0.43
• Roof - Pitched 300mm insulation - U-Value 0.12
• Windows - wood/UPVC frame, Triple Low-E - U-Value 1.70
• Floor - concrete foundation no insulation - U-Value 0.70

So heat loss will be overestimated at least a little. There will be ~2kWh/d loss from the DHW tank into the kitchen also, given the tank manufacturer's specifications. I asked if a recalculation can be done based on my U-values, to see if the upgrade of the hall and kitchen radiators can be avoided.

2024-09-26: Recalculating...

Octopus got as close as possible to those with their software as I understand it, and got to a design-day heat loss of 2104W. (Walls were dropped to 0.27 and floor to 0.24.) That roughly halves calculated design-day demand from 3805W.

Thus now only one radiator replacement is required — in the living room. And I suspect that annual heat demand will be much closer to measured actuals.

I pulled the remaining Octopus actions from above into an email and received a comprehensive response:

Have Octopus confirm that its installers will leave cabling and pipework easy to add electricity and heat metering.

Our engineers can leave accessible pipework and cabling for the customer, this would need to be discussed with the engineers directly.

Have Octopus confirm that a Cosy 6 unit will be available if required.

This would require a resurvey to confirm this and depend on what requirement you mean - we would look to go ahead with the current Daikin model as per our current plans ideally.

Have Octopus confirmed that its installers will leave pipework easy to reconnect the Thermino and Intasol per the second configuration option.

This is potentially possible but will also need to be discussed with the engineers either at a pre-install visit or during install.

Have Octopus ensure that buffer and expansion tanks brought to site are no larger than the indicated 300mm diameter, and that the DHW tank is slimline and no larger than 475mm diameter (smaller would make life easier), and that placement of all the blue Octopus-fitted items as shown in the "arrangement 3" plan and side views is acceptable, not just the DHW tank.

The equipment we provide will be of those diameters, as for the arrangement, a PIV [Pre-Install Visit] would be best so you can confirm this with the installer.

Ask Octopus for more clarity on the Daikin 'app': in particular can either heat pump be controlled without a smartphone, and what happens if (say) Daikin goes bust or my Internet connection is down?

The customer will have full control of the heat pump via the MMI (user interface) which will be fitted near the hot water cylinder, this does not require the internet to work.

Ask Octopus if any basic consumption data available from either heat pump without installing a CT clamp?

The MMI can give basic run time and energy produced in kWh, but this is not real time data and does not produce a COP/SCOP.

Ask Octopus for more clarity on the warranty issue: why would fitting heat metering violate any warranty compared with (say) replacing a TRV or fitting a decorative higher-output radiator upstairs, or reconnecting the DHW heat battery, for example?

If any changes post install by the customer/third party to the system causes a fault with the heat pump this may not be covered, any changes which affect the performance of the heat pump won't be covered.

Ask Octopus to explicitly confirm on the warranty DHW issue: that recommissioning the solar/heat-battery part of our system will not not violate warranty.

Again, any changes post install by the customer/third party to the system causes a fault with the heat pump this may not be covered, any changes which affect the performance of the heat pump won't be covered.

Ask Octopus for more clarity on immersion control: in particular can it be controlled in a grid-friendly low-carbon way such as from my spare Eddi channel?

The immersion on the hot water cylinder can be connected to an Eddi, we will need to be aware of this pre-install to advise the team accordingly and check viability.

Ask Octopus if it can be made easy to place one or more additional temperature sensors on the DHW tank.

The cylinder we provide only has 1 sensor pocket for this.

Stress to Octopus need to be especially careful to avoid decor damage on radiator refits.

This is worth mentioning at a pre-install visit with the installer direct but I can make a note on the install booking to be especially careful.

Ask Octopus what colour finishes are available for the Cosy 6 as blue is not acceptable.

The Cosy also comes in a dark grey colour.

Octopus also described how its normally allows 3 to 5 days for an installation, with a schedule something like:

• Day 1: remove old boiler, install new DHW: heating and DHW off.
• Day 2 to 4: pipework, radiators and heat pump installed: heating off, DHW on.
• Day 4 to 5: full commissioning: everything back on.

We are aiming to get the heat pump installed before December so that we don't need to do any space heating with gas this winter at 16WW. We are in the process of setting up a pre-installation visit for this month if possible.

The Octopus crew can also remove our gas meter during the process and Octopus said:

We can remove the meter but you'll need to contact [Ecotricity] to remove the standing charge. They sometimes request a document confirming we have carried out the work before they cancel the charge, but in most instances this is not necessary (we can issue this document on request regardless).

Install Date

The next day a provisional install date near the end of November was offered, before we normally turn our heating on.

And the following day the date was confirmed.

I have let our retailer (Ecotricity) know the plan by a Friday-evening email, with the expectation of moving to an electricity-only tariff in December. On Monday morning Ecotricity helpfully replied with:

Thanks for the update.

Please provide us with photos confirming octopus removed the gas meter during the heat pump install once complete.

I cannot close your gas account until we receive the evidence. Once we have the evidence, I will need to arrange for your gas meter to be removed from industry manually. This can take up to 10 working days to complete.

Note, your standing charge will stop from the date of removal.

Today I have received an email about Giving consent and confirming your eligibility to have Octopus apply for the BUS grant on my behalf.

I have to fill out a form on-line amongst other things confirming that I do and will (co-)own the property, live there, have a recent EPC, my property has not previously had a heat pump or biomass boiler that was funded by the Government or by the Energy Company Obligation (ECO), ... the details I've confirmed on this application, as well as the new heating system at my property, may be subject to audit ....

I called our home and contents insurer to let it know the due date of installation, ie at which we would be on an ASHP. All it seems to care about is that we are not using a single-skinned plastic oil tank out of warranty!

There does not seem to be any premium change required.

As of the start of this month on our Ecotricity dual-fuel tariff (Green Electricity and Gas), the electricity to gas kWh ratio is just over 4 (28.78p vs 6.74p ex-VAT, so 4.3). With the new heat pump using weather compensation, and the positive interaction with my Radbot TRVs per [hart-davis2024zone], and for separate reasons DHW also because of current heat battery use, I believe that our OPEX will fall marginally compared to being dual-fuel. Our carbon footprint should fall substantially. Most of our annual gas bill is the standing charge at over £100.

Dale Vince, founder of Ecotricity, remains against heat pumps and would prefer home heating with bio-gas from repurposed UK farmland [rose2021grass] [savage2024backlash].

I received an email apparently from Ofgem saying that We have been unable to complete our verification checks using Experian with the information provided to us in the application provided by your selected installer and asking me to send photo ID and proof of address to another apparent Ofgem email address.

This could be a clever phishing attack, but I do not think that it is.

I think that it is unwise to send especially the ID by email. I called the listed helpline and the chap on the other end agreed. I am hopeful that this can be resolved quickly.

2024-10-25: alternatives

It seems that I am going to be be given an alternative way to get ID to Ofgem.

2024-10-30: approved

This evening I received an email from Ofgem: We have approved Octopus Energy Services Limited Boiler Upgrade Scheme (BUS) application for installation of Air source heat pump...

The voucher expires 2025-01-30.

Today there has been a pre-installation visit (PIV) by Octopus, to establish details of the works that I need to get done before the main installation crow rocks up next month. (Photos.)

I have learned a lot today, and we have agreed how the external unit should be positioned and where the key pipework and cabling should go, options for the condensate run-off handling, and how to leave internal pipework to make it easy for me to reconnect the heat battery first and heat metering later.

The engineer who came today cannot make firm promises on behalf of the installation crew, but what I note below is what we think will work (preferred option) and some alternatives.

External unit positioning and base

There is a platform ~1200mm wide in front of the kitchen window with its left edge (looking from the house) at a flower bed 1000mm from the property boundary, and with its right edge at the footpath to the front door (see photo). It is covered with 300mm square tiles, thus 4 across its width. The row of tiles nearest the house is slightly less than 300mm front to back.

For Octopus I need to prepare [a 1500mm x 600mm flat and level concrete base to take the ~100kg weight of the heat pump external unit] enough space for a few centimetres of gravel as soakaway under the unit. That corresponds to 5 x 2 tiles.

There is one (curved) corner of the Daikin unit that has the evaporator coils and is the most fragile. Our plan is to have that in the bottom left out of the way, one tile in from the bottom-left corner of the platform, with the unit running long ways away from the house. The badged front of the unit will point towards our porch. This keeps all of the unit more than the required 1 metre from the boundary, while minimising visibility from the road, protecting the most the vulnerable part of the unit, and allowing enough clearance all round for servicing.

Octopus need the concrete base only built flat and level; I am not meant to be embedding any bolts nor making any holes in it.

I intend to put up some screening / fencing so that the unit is barely visible from the road and footpath (and car park) as far as is reasonably possible.

External water pipework and trenching

Given the unit orientation as above, water and power will run from the left (rear) of the unit at the far end from the house (see photo).

The flow/return water pipes are run together in a plastic 4" (~10cm) outer. The mains power will need to be ~16A and will run via an isolator on the wall to a secondary consumer unit and then into the main consumer unit. There will be control wiring (but no power) from the external unit to the internal part of the system.

These will all run initially in a trench from the rear of the external unit to the corner of the house, to the left of the platform as seen from the house.

The trench needs to be at least 750mm deep by 150mm wide, with no sharp bends.

The plan is to make the trench the entire width of the left-hand row of 6 tiles next to the flowerbed, the tile below the platform immediately in front of that row, then some or all of the nine tiles between the flowerbed and the house, ie to the right of the drain in the photo where the drain is visible. It is not yet known where Octopus will core-drill the entry hole for the water pipes and signal wires.

When drilling Octopus will have to be aware that the aerogel IWI may snag the bits, and airtightness will have to be restored. This will also apply to the penetration for power for the external unit.

External unit power

I assume that Octopus will fit the external disconnect roughly above the house end of the trench, then route a cable at ground level or above the window to the door, make a penetration close to the door, and run cable into a small new consumer unit with RCD. That will then wire into the spare way in the main consumer unit.

I would like to have a current clamp on that sub-consumer unit to measure total heat pump electricity consumption at the H4 boundary. (This will miss consumption of the MMI if it is powered from the existing combi's spur as now seems likely.)

I have today ordered a new CT clamp for the Eddi at just under £21 including delivery and VAT, which I would like fitted onto the cable when the sub-board goes in. The end at the Eddi can be connected while Octopus is on site or later if need be.

Action (CT-clamp-fit): ask Octopus to do a CT-clamp-fit while providing power to the external unit.

Condensate handling

The external unit will generate condensate run-off, particularly on cool (above freezing) humid days. This liquid water extracted from the air has to be disposed of. Given the positioning and orientation of the unit suggested above, the condensate will emerge on the end of the external unit nearest the house, on the left-hand side ie the rear of the unit.

There are three plausible solutions for the condensate that we discussed:

1. Soakway 1: lift the partial tile at the edge of the platform between that corner of the external unit and the house; dig to the same depth as the trench next to it; fill with gravel and sand and run a very short pipe into it.
2. Soakway 2: excavate the flowerbed behind the unit, fill with gravel and sand and run a pipe into it; put the soil back on top to allow flowers to be planted again.
3. Run a pipe to the external drain at the front of the house.

Immersion heater and pasteurisation cycle

I would like to power the DHW immersion heater via the Eddi boost circuit 1 currently used for the Thermino.

I would also like to install a connection from Eddi boost circuit 2 to power the Thermino heater as the 'overflow' storage once the DHW cylinder is 'full' of heat.

This means that the power available at the existing combi spur should suffice for the MMI and any other local controls.

Internal elements positioning

The Octopus engineer seemed happy with the side and plan views (arrangement 3), ie that there would be sufficient space to fit things in.

Internal pipework

Pipes and control wiring from the front corner of the house should run under the kitchen units on that (north) side to where the DHW tank etc will be installed.

I will remove boxing etc before Octopus arrives, and reinstate afterwards as needed.

The Octopus engineer said that he will talk to the installation crew and ask them to provide teed-off sections in the DHW side hot and cold runs so that both can be easily be re-routed via the Intasol, and will also ask them to leave ~500mm runs in the main flow and return with isolation valves on each side to make it easy to insert heat flow devices.

Control unit (MMI) power and abilities

The MMI etc can probably can be powered from the spur that currently runs the combi since it will not need to power the immersion heater.

The MMI and the system generally does not need a live Internet connection in order to function, nor a smartphone app nor other fluff.

Apparently it is possible to set a DHW temperature/time profile in the MMI, eg low from 4pm to 7pm to be grid friendly as described previously. I may wish to take over finer control later but this is good for now.

Apparently I can call for space heat with volt-free contacts (nominally from our house thermostat, in practice from the OpenTRV controller), and while flow temperature will be driven by weather compensation I can have a setback (or entirely off) overnight to avoid sleep disturbance.

2024-10-25: proposed static schedules

• Central heating on (with Radbot call for heat) 07:36 to 15:54 and 19:06 to 21:29, and off or with a heavily-set-back flow/target temperature outside those times. Possibly a raised flow/target temperature 15:06 to 15:54 per [berry2023flexibility].
• DHW default temperature 50°C when heating is on (07:36 to 15:54 and 19:06 to 21:29), 45°C outside those times. Possibly a raised target temperature 55°C (max from heat pump) 15:06 to 15:54 and lowered 40°C 15:54 to 19:06.

Outside waking hours ~07:00 to ~21:59, or to match with set-back periods as above, the heat-pump should run in a quiet mode if at all.

Morning start time and 4pm to 7pm off time is chosen to minimise draw at GB peak electricity/heat demand and peak heat demand ramp times [watson2019peak].

All times local, not UTC. Start times in particular should avoid exact half-hour boundaries (and exact 5-minute boundaries and/or 5 minute gaps from HH boundaries) so as to be more grid friendly and avoid "herding" with other precision-timed grid users.

We had a second pre-installation visit (PIV) today, in part so that the Octopus engineer could talk to our builder (DM) about externals. There will be a bit less work to do outside than we had anticipated. DM will come once to do externals and once to do internals.

Probably the most interesting bit will be getting power out from our metal-boxed very busy electrical incomer board to a new consumer unit. Octopus avoids routing power via any existing consumer unit. But we have done stuff like this before, and maybe we have a bit of a plan.

It seems that the power to the in-house MMI, diverter, etc, comes via the cabling and pipework brought in through the wall, so a current clamp at the new Octopus consumer unit will be a true H4 boundary measurement, hurrah!

MMI

I have been sent the MMI 'basics' documentation to look at. I think that I will record what settings Octopus choose, and then add complexity gradually as I get to understand the system. In part these will be putting in sensible 'static' temperatures and timings, with a hope of being able to dynamically override them later, eg diverting PV generation to DHW with the heat pump rather than immersion.

I have set up extra energy variable monitoring to cover the monitoring of the heat-pump electricity consumption (H4 boundary) and the second Eddi heater output to drive the heat battery.

I assume that if the DHW tank is heated enough that the immersion heater thermostat cuts out then the tank is sterilised or will be shortly. For this the immersion thermostat temperature should be set to >60°C.

An initial simple zero-software pasteurisation control can be to program the Eddi heater 1 boost via the Eddi front panel for ~4h each Sunday from ~2am (nominal ~8kWh maximum input with 2kW grid limit), which should be enough to heat the DHW tank to >60°C and force the DHW immersion thermostat open. Grid demand and carbon intensity are both likely relatively low then.

A more sophisticated approach would be to wait until a 'supergreen' time in the small hours preferably on a weekend around 30 days after the heater 1 thermostat was last open (ie up to temperature, >60°C) and boost heater 1 (DHW tank) until the thermostat opens again. This could all be done on the heat battery RPi, eg the supergreen flag can be accessed with wget.

It may be possible to detect the DHW tank hot and immersion thermostat open with some combination of "sta" (status) 5, failing to put energy into heater 1 when trying to boost it, or sending any energy to heater 2 having implicitly failed over eg during diversion.

In preparation for the heat pump install later, and the disconnection of the Thermino heat battery well ahead of that so that Octopus' install is as vanilla as is reasonably possible, I have replaced the heat battery target script with a stripped down version that always sets a maximum top-up target of zero. That way no more boost energy will be absorbed and the Thermino will be cold before disconnection time.

The heat battery is currently about half full. The next couple of days are forecast to have quite high grid carbon intensity.

2024-11-15: Briefly back on

I am briefly re-enabling boost overnight Friday to Sunday morning, so as to capture a battery-full of low-carbon grid energy for my weekend hot bath and other uses, given the grid intensity forecast, and that no interior prep works were done today.

DM and assistant arrived before 8am. It is cool (~10°C) and overcast. Exterior ground prep works finished before noon.

It is frosty outside this morning (and will likely hit winter design temperature ~-2°C over the next 48h) but we are resisting putting on the gas central heating. It is ~14°C inside.