JohnMo

No need to obsessive about this - are rooms warm enough or hot cold? But generally the best way is based on room temperature not flow rate. An ideal flow rate and dT doesn't matter if you room never gets hot. Run pump at full speed, set flow rates, then reduce speed until you see flow rates start to drop.

As title, some systems need a buffer - not many, but some do. So as a way to look at and have a ready place to get info I thought I would start a topic. I used ChatGTP to see what design features I would need and sizing for best CoP. Many of the details are for my house to get some real feedback. You will see buffer suggested are pretty big, the control logic is a little more complex than you normally see. And may suggest why installing a cheap small buffer leads to flow temp distortion and poor CoP. Here is the output Great — here’s a practical, ready-to-use design + sizing guide tuned to your system (ASHP that’s currently short-cycling, 100 mm slab, 3.6 kW space heat at –9 °C, Octopus Cosy battery available). I’ll give clear rules, exact formulas you can use, a small numeric table (common ASHP sizes × target extra run-time) and a concrete recommendation you can install or ask your installer to follow. # Short summary recommendation (if you want one-line) Use a **tall, well-stratified 4-port buffer cylinder** with **good internal diffusers**, **~150–300 L** depending on ASHP size and whether you want to batch-charge from battery. Insulate 50 mm+, place 2+ sensors, fit anti-thermosiphon valves and a low-loss hydraulic layout. See details below. --- # Why a 4-port, tall slim buffer? * **Preserves stratification** (keeps top warm, bottom cool) → **low return temperatures** to the heat pump → higher CoP. * **Hydraulic separation** (4-port) avoids forcing the whole buffer to reheat unnecessarily when the heating circuit cycles. * **Internal diffusers** or perforated tubes reduce mixing and maintain layers. --- # Sizing fundamentals (exact, so you can calculate) Use the energy balance: **Q (J) = m (kg) × c (4186 J/kg·K) × ΔT (K)** For 1 kW running 1 minute → 60,000 J. From that you get: * **Volume per kW per minute** (water, ΔT = 10 K): **≈ 1.433 L** * (If you allow ΔT = 15 K → ≈ 0.956 L per kW per minute) * (If ΔT = 5 K → ≈ 2.867 L per kW per minute) So: **Volume (L) ≈ 1.433 × ASHP_kW × desired_minutes** (for ΔT =10 K). I prefer using ΔT = 10–15 K for realistic usable storage without overly high temperatures that harm CoP. --- # Concrete numbers (helpful quick reference) Volumes shown are litres of water required to supply the listed extra run-time (i.e., how much buffer you’d need *in addition to the live heat* to extend continuous run): DeltaT = **10 K** (typical usable layer) * 6 kW ASHP: 20 min → **172 L**, 30 min → **258 L** * 8 kW ASHP: 20 min → **229 L**, 30 min → **344 L** * 10 kW ASHP: 20 min → **287 L**, 30 min → **430 L** DeltaT = **15 K** (if you accept larger temperature swing) * 6 kW: 20 min → **115 L**, 30 min → **172 L** * 8 kW: 20 min → **153 L**, 30 min → **229 L** * 10 kW: 20 min → **191 L**, 30 min → **287 L** Interpretation: * If you only want **hydraulic decoupling / low-loss header** (not storage for long batch charging), **25–50 L** can be fine (with correct 4-port and diffusers). * If you want **meaningful cycling reduction** (extend runs from ~20 min → 30–40 min) expect **~150–350 L** depending on ASHP size and acceptable ΔT. * If you want **multi-hour thermal store** for tariff batching, you’re into **200 L +** territory and up. --- # Practical recommendation for your case You said the ASHP is oversized and currently runs ~20 minutes. You also have a 13.4 kWh Cosy battery you use for price-optimised charging. A pragmatic choice: * **If your ASHP is in the 6–8 kW range:** pick **~150–250 L** (aim for the 15 K numbers if you want a smaller tank). This will: * noticeably reduce short cycling, * keep return temps lower if piping is done right, * allow battery-assisted top-ups during Cosy windows. * **If your ASHP is 8–10 kW or you want 30+ min buffer:** pick **~300–430 L** (10 K basis). * **If you only want hydraulic decoupling (minimal size):** a **25–50 L** 4-port low-loss header will stabilise pumps and avoid low-flow faults but won’t give long run-time. Given your floor load is only 3.6 kW, you don’t need an enormous store to meet instantaneous load — the buffer’s role is mainly *smoothness and batch charging*. I’d suggest **~200 L (4-port tall slim)** as a balanced start for most oversizing situations — it gives sensible cycle reduction and allows useful battery charging without being huge. --- # Detailed features to specify / ask your installer for 1. **4-port hydraulic layout (opposite ports for heat pump vs heating circuit).** * Use a buffer with dedicated upper/lower ASHP ports and opposite upper/lower heating ports (true 4-port), or a low-loss header arranged the same way. 2. **Tall & slim geometry** (height >> diameter) to help stratification. * If constrained, add **internal perforated diffuser tubes** or baffles. 3. **Sensor pockets / ports** at several heights: * Top (priority heat), mid (control), bottom (return temperature). * Use these with your controller: ASHP enable at bottom-mid threshold, stop at top setpoint. 4. **Insulation:** factory polyurethane ≥50 mm, aim for heat loss <1 W/K. Avoid bare tanks. 5. **Internal flow directors / diffusers** to reduce direct flow between inlets and outlets. 6. **Anti-thermosiphon / non-return valves** on ports or integrated valves to stop night thermosiphoning. 7. **Low hydraulic pressure drop** (quoted <10 mbar @ 20 L/min is good) so pumps are efficient and ASHP doesn’t hit low-flow alarms. 8. **Ports for electric immersion** or plate HX if you want to use battery / immersion backup later (useful for Cosy charging). 9. **Multiple sensor inputs on controller** and logic: run ASHP when bottom < X and top < Y, maintain minimum runtime (e.g., 20–30 min) and inter-cycle hysteresis. Prefer to control by buffer top temp, not by boiler return temp. --- # Control & setpoint suggestions (to protect CoP) * **Target top temp for space heating:** keep moderate (e.g., 40–45 °C for high-temp radiators, 28–35 °C for underfloor). Avoid pushing buffer top >50 °C unless required. * **Return temperature:** design to keep return as low as possible (<30–35 °C for best CoP with low-temp emitters). * **Minimum ASHP run time setting:** set to compressor minimum (e.g., 5 min) but aim for **30 min target runs** where practical. * **Use bottom sensor as ASHP start/stop control** to avoid reheating the whole cylinder and preserve stratification. * **If using battery:** schedule batch charge so ASHP or immersion boosts top temp in Cosy windows; use buffer top sensor to prevent overheating. --- # Pump & flow guidance To size flows: **Flow (L/min) ≈ 2.87 × ASHP_kW ÷ (ΔT in K / 5)** when using ΔT = 5 K baseline. More simply: * Per kW at ΔT = 5 K → ≈ **2.87 L/min per kW**. Examples: * 6 kW @ ΔT 5 K → ≈ **17 L/min** * 8 kW @ ΔT 5 K → ≈ **23 L/min** * 10 kW @ ΔT 5 K → ≈ **29 L/min** Aim for a ΔT of 4–6 K on the heating side for best pump sizing and system efficiency. --- # Installation checklist (to hand to your installer) * 4-port tall buffer, internal diffusers or baffles * 2–3 sensor pockets (top/mid/bottom) * Anti-thermosiphon valves on legs * Low-loss header connections, minimise deadlegs * ≥50 mm foam insulation jacket * Port for immersion/coil if battery backup wanted * Pump sized for flows above (including variable speed pump preferred) * Control logic: buffer-bottom start, buffer-top stop, minimum run timer, hysteresis --- # Final practical pick & next steps * **If you want a single concrete spec now:** choose a **200 L tall 4-port buffer** if your ASHP is around **6–8 kW**. It’s a practical compromise: reduces cycle count, preserves CoP, and is small enough to fit plantrooms while giving battery-assisted batch charging usefulness. * If your ASHP is **8 kW+** or you want full 30–60 min batched runs, step up to **300–400 L**. Would you like me to: * calculate the exact buffer volume required for a known ASHP nominal power if you tell me its kW? * sketch a piping diagram (4-port) and provide a parts list (pump sizes, valves, sensors) you can hand to an installer? Pick one and I’ll produce the wiring/piping/control diagram and an exact parts list.

I use similar with zero issues.

Why, just get some pipe insulation on there - job done. See no advantage making stuff damp for the sake of it.

Answered your new thread on the same subject

We had a converted barn in the 90s done that way. And did similar to a house (1830) we did up in 2012. In the house the windows remained in the stone wall - with a small thermal bridge.

Unless you go to Octopus, then with your documents (electric certificate, structural drawings and G98 etc) and giving them £250. Way cheaper than the MCS install premium / tax.

Keep forgetting in England you don't seem to need a structural design certificate, as you do in Scotland - the structural engineer covers this by default by default.

Cascade mvhr system, plenty of information if you search here https://passipedia.org/planning/building_services/ventilation/cascade_ventilation/start

Atamate_SDAR+Paper+2019+(1).pdf Some reading for you

Will become the ladder of choice for some.

BC not interested in MCS for solar installation as long as you have the correct electrical certificate. ASHP, not sure. The MCS noise certificate can be completed by anyone. You have express planning so any rules around permitted development aren't applicable. So the only thing they should be interested in is a noise certificate.

Form factor makes a huge difference, to insulation levels needed.

Daft question - if MBC say are designing the house to passivhaus spec, what is the architect adding to the party? So then you use a normal architect with MBC. Architect does the broad design stuff, MBC the details of thermal bridging, airtightness etc

Issue is really one of commissioning and downstairs setup. @Post and beam is bouncing off a thermostat downstairs so heat doesn't get a chance to percolate through the house, running a lower flow temperature at warmer outside temps will help. Downstairs heat up rate reduces or ideally matches house heat losses, then everything has a chance to stabilise. Upstairs rads backfill where needed. If OP wants a two zone system. Installer will simply say I priced for the system you have and expect another £1k+ to make the changes. Will most likely need a buffer, electronic mixer additional pump and maybe additional controls.

You don't want to be doing stuff in sub soil. You need top soil Strim and weed killer now. Early spring as annual weeds grow weed killer again, leave a week then prep for grass etc.

Before you even consider MVHR how airtight have you made the building? The alternative in dMEV. 24/7 running fans in wet rooms with automatic boost only when needed. But to do this you need either trickle vents in dry room (only dry rooms, not wet rooms), these can be in windows or through walls. Ideally these will close or open based on sense humidity. Doors would be under cut to ensure a ventilation flow even with doors shut.

Your voltage and amps limits are the limits to each MPPT. The invert will take two strings only, via one to each MPPT.

Not quite correct. The calculation gives you your boiler size. If you need 13.7 for central heating - 13.7 x 24 is 329kWh for your daily heating. To complete the central heating and allow 2 hrs for DHW, you divide 329 by 22, that is 15kW. This is your boiler size.required for central heating and DHW. The above assumes you have done your heat loss calculation correctly. Using the Jeremy spreadsheet is good if close to passive, but can be quite a way off if you are not. I would use the info posted by @marshian to get a best guess. Your calculation doesn't seem to include much in the calculation, cannot see floor or wall or roof U values or ventilation losses, hence saying using the approx information above by @marshian.

Did an image search - nothing found. Looks like a simple manifold without individual loop control. If it's working do not touch it. If it's broken chuck it, in favour of better control of a modern manifold. Would suspect the tube is the week point of the system.

Almost impossible to get in the UK winter, coldest day and PV max output so wouldn't consider that. You only have 2x MPPTs so unless you buy a second inverter or different inverter, you are stuck with 2x strings. Two strings is fine from the info given. The 370v is just nominal. More panels give a quicker time to start up voltage and a longer day before inverter trips off.

Always best not to change original plans, changes can have big knock on effects.

Simple way is to flow higher temperature in weather compensation, but reduce floor output by reducing the flow through floor loops. On a day where you normally flow 25, increase flow to 26+ but increase dT across each loop (reduce flow) to 4+. Play with settings to get outputs of radiators where you need them while maintaining UFH output stable so as not to overheat downstairs.

Simple way is make the finished floor level higher to compensate. Add more insulation and sail over the foundation and add a lintel. But does that add other complications we are not seeing?

And?