Earth Notes: PhD Research Zoned Heat Pump Modelling
Updated 2025-04-22.Overview
This CS1 research strand is attempting to test the assertion that TRVs and microzoning are a relatively inexpensive way of enhancing energy (and carbon footprint) savings and comfort in UK dwellings retrofitted from gas-fired heating with radiators to ASHP. This is also looking out for other pros and cons such as increased wear from cycling.
My home heat pump system control is intermittent, with zoning, and some setbacks and grid friendly aspects.
Overall Study Question
Does TRV microzoning of domestic space heating in the UK with retrofitted ASHPs save electricity and carbon emissions and money overall, and improve comfort?
Wishlist/Plan
An overall wishlist/plan (updated as needed):
Develop static EnergyPlus model and heat-pump/TRV behaviour for bungalow/detached
Recreate [hart-davis2024zone] bungalow and detached house in EnergyPlus with a plausible heat pump heat generator using pure weather compensation. Note that the setback case corresponds to permanently turning down 50% of the heated space, and the non-setback case to turning down 0%. Build measures of energy use and comfort at 0% with and without buffer tank / volumiser, and at 50% with buffer tank / volumiser, ie three base scenarios. The buffer tank / volumiser is sized to meet heat pump minimum volume requirements. A possible separate bypass may be needed to ensure minimum flow requirements.
The three base scenarios are a static equilibrium state at design conditions.
The notion of comfort, possibly based on EnergyPlus available metrics, will need to be developed.
The heat-pump element will need to be able to be plugged into the Energy House 1 EnergyPlus model that has been made available.
Dynamic Scenarios in bungalow/detached
Extend the model to work against the weather tapes used in [hart-davis2024zone] or similar, ie time varying. Depending on details of the EnergyPlus implementation, eg thermal capacitance, results may diverge somewhat from [hart-davis2024zone].
Consider the effect on energy and comfort of combinations of zero or more of:
- Adjust the TRV behaviour to be time-dependent rather than constant, eg from scheduled/smart TRVs, eg to turn down different areas at different times of day (eg bedrooms during the day), to compare against an equivalent overall/mean percentage turned down across each day.
- Set back the heat pump flow temperature overnight a few degrees.
Energy House 1 as validation?
Replace the current gas boiler heat source in the extant Energy House 1 EnergyPlus model with the heat pump model from above, possibly scaled for relative heat loss, with no zoning. Run at constant design day conditions and then with some weather tapes analogous to those from [hart-davis2024zone]. Take these as base cases.
Test with zero or more combinations of the features from above:
- Buffer tank / volumiser and/or bypass valve, as would be needed to meet heat pump requirements for the most closed-down zoned implementation.
- 50% or 25% of the heated space statically and dynamically (when using weather tapes) zoned with TRVs.
- Heat pump flow temperature setback overnight (when using weather tapes).
- CIBSE/SAP working-family occupancy profiles or other occupancy profiles (when using weather tapes).
Part of this is to see if potential energy and comfort improvements can be adequately modelled with static (no diurnal variance) conditions and static TRV setbacks, which would markedly simplify the physical verification element if so.
One or two of these scenarios would then be validated in Energy House 1.
2025-02-24: Replicating the Original Scenario
Aim
Build an E+model with the HG bungalow initial (design) conditions, but as 1 open space at target (non-set-back) temperature, and model matching expected heat flows from a heating system into the inside (somehow) to the outside.
2025-04-13: update
Though I have not yet found a way to turn off all the cleverness in EnergyPlus, I now seem to have an OK model ("bungalow-1
" as of today, in V0.0.3
, https://github.com/DamonHD/BuildingModelsForHeating/tree/main/ZonedHP/bungalow-1/20250413-snapshot) which seems to be within a few % of the initial 'design day' scenario and a full year simulation. I think that I can use that as the basis of the next part, working on the internal partitioning and heat flows with setbacks in ABAB and AABB configurations, and THEN some initial heat pump 'bad setback effect' work.
This does not preclude coming back and improving this base model and rebuilding the partitions etc on top of it.
Possible next steps:
- Properly calibrate
bungalow-1
heat demand against paper initial -3°C/AAAA scenario, then against (say) London and Birmingham with weather/temperature data. - Create internal partitions, ie 4 rooms/zones, to make
bungalow-2
. - Cross check
bungalow-2
heat demand against paper in AAAA, ABAB and AABB configurations, initial and (say) London and Birmingham.
Further out:
- Swap the current heat source for a simple heat pump and zoning:
bungalow-3
. - Upgrade the heat source to enough fidelity to have varying CoP, cycling, etc, though not necessarily any manufacturer's spec:
bungalow-4
. - Match heat pump behaviour more closely to one or two manufacturers' unit specs/curves such as DHD's Daikin 3 M, the Vaillant proposed for EH1, etc, with volumiser and other key elements in the model.
2025-04-22: EH1 Proposed Combinations
Things to test in EH1:
- Volumiser efficiency and wear effects. Volumiser (and bypass) must be in-circuit when any TRVs are.
- TRV/zoning efficiency and comfort and wear effects.
- Steady external temperatures, insolation, etc, vs diurnal day/night pattern, looking for efficiency and comfort and and wear other (eg model accuracy) effects.
- Setback efficiency and comfort and wear effects.
- Self-balancing TRV/zoning efficiency and comfort and wear effects.
Proposed method
Over the course of 2 or 3 weeks at Energy House 1, with external temperature and other conditions corresponding to a typical winter day in the UK with, test the first 5 or all scenarios in the table below.
- External temperature steady –3°C, or with typical UK diurnal cycle and mean –3°C.
- Interior spaces normally targeted at 21°C, initially statically balanced with fully open circuit and steady external temperature and flow temperature selected to match the heat loss and no volumiser in circuit.
- For zones that are set back (25% or more of internal area, eg small bedroom, hall/stairs, kitchen), target temperature is 18°C and achieved with TRVs in these zones and TRVs set to or just above 21°C in non-setback zones. The volumiser (picked to ensure minimum system volume is available even if all TRVs closed) must be in circuit and the bypass valve available.
The primary reporting metrics from all scenarios would be:
- Total electricity consumption by the AHSP (at H4, money and carbon costs).
- Total heat output of ASHP unit (COPH4).
- Over- and under- heating levels and durations as an indication of (dis)comfort.
- Other effects good and bad such as cycling count for the ASHP (contributing to Total Cost of Ownership).
Scenarios
Scenario Name | External Conditions Diurnal | Volumiser | TRVs | Setback In Some Zones | Min Run | Effects Measured |
---|---|---|---|---|---|---|
No, Yes | No, Yes | No (open), Simple, Balancing | No, Yes | H | ||
Control | 24 | Simplest base case for installation/setup, and eliminating effects of thermal capacitance. | ||||
Volumiser | Yes | 24 | Overall energy (and pump wear) effects of volumiser, vs Control. | |||
Diurnal-Control | Yes | 48 | Comfort (temperature stability), also pump wear. Diurnal energy base case, vs Control. | |||
TRVs-no-setback | Yes | Simple | 24 | Comfort and overall energy (and pump wear) effects of TRVs (set to or just above target temperature) on top of volumiser effects. | ||
TRVs-with-setback | Yes | Simple | Yes | 24 | Comfort and overall energy (and pump wear) effects of TRVs (mainly set to or just above target temperature, lower in some areas) on top of volumiser effects, vs Volumiser. | |
Diurnal-TRVs-with-setback | Yes | Yes | Simple | Yes | 48 | Comfort, overall energy, pump wear vs TRVs-with-setback. |
SB-with-setback | Yes | Balancing | Yes | 24 | Comfort and overall energy (and pump wear) effects of self-balancing TRVs above simple TRVs. | |
Diurnal-SB-with-setback | Yes | Yes | Balancing | Yes | 48 | Comfort, overall energy, pump wear vs Diurnal-TRVs-with-setback. |
IN PROGRESS
References
- [hart-davis2023waste] Do Thermostatic Radiator Valves Waste Energy in UK Heat Pump Retrofits? (PREPRINT)
- [hart-davis2024zone] To Zone or Not to Zone When Upgrading a Wet Heating System from Gas to Heat Pump for Maximum Climate Impact: A UK View
(Count: 2)