Daikin Pre-plumbed Cylinder (WorldHeat) Buffer Strangeness

Hardware Heatpump
,
1

I have been looking at the water flow and return temperatures on the 4-pipe buffer which forms part of the Manchester-made WorldHeat pre-plumbed cylinder which is installed with a Daikin EDLA. Something is quite wrong so I am seeking other people with the same hardware to “compare notes” with a view to finding out whether there is something wrong with the install, manufacture of my tank, or the fundamental buffer design in these units.

The buffer appears to be plumbed in correctly. Reading from left to right around the base, where the pipes exit, the order is Heat Source Return, Space Heating Return, Space Heating Flow, Heat Pump Flow. This matches the picture in the installation manual. The body of the secondary circulation pump has an arrow cast into it showing flow is indeed into the SH return. When I got the cylinder, however, the pump was on back-to-front (pumping water out of the SH return pipe!) but I spotted this and got the installer to reverse it.

The issue is that, having monitored the water temperatures (from the outside using DS18B20 devices cable-tied to the pipe, with some thermal paste and surrounded by several turns of fleece fabric for insulation), I can see the water which leaves the SH flow is the same temperature as the HP return. A typical approximate reading for “heating on” would be HP flow 32C, HP return and SH flow 28C, SH return 25C. This is clearly BAD; I’m paying a COP penalty for heating the water to 32C whereas heating to 28C would be sufficient for the same space heating, given an ideal buffer tank. This situation persists for a range of pump speeds (my original plan was to monitor the temperatures so that I could modulate the secondary pump to properly balance the buffer tank).

It looks like either the HP or SH pipes are transposed.

SO… I am looking for evidence:
A. Does my arrangement of 4 pipes exiting the buffer match yours?
B. Which direction is your secondary (space heating) circulation pump (arrow on metal body)?
C. Do you have any temperature readings? All 4 pipes at the buffer would be best (either using a similar sensor setup to me, or maybe an infra-red remote temp meter would do). Alternatively, do you have evidence that the space heating flow temperature is close to, or far from, the leaving water temperature reported by the heat pump?

In spite of this issue, the heat pump appears to be behaving itself, with longish periods of steady energy input (SMD120 feeding my Open Energy Monitor). Its attempt to maintain its flow-return dT is, of course, unhelpfully suppressing my emitter temperature!

Thanks in advance, Adam

2

Not sure about the design but I changed my 4 pipe buffer into a three pipe one - videos and text out there - improved performance no end.

3

Hi Adam.

I don’t have the WorldHeat buffer like yours, just a simple vertical 4-nozzle 50 litre buffer tank that my installer chose (heat pump in/out and emitter out/in connections with a secondary fixed-speed pump on the emitter circuit).

I installed decent quality surface T/Cs on the four inlet/outlet lines (with ht cement and plenty of insulation) some time ago, and more recently I had a Sontex 389 heat meter installed on the emitter side of the buffer tank, so I could see what energy was actually entering the house (as opposed the the heat supposedly leaving the Outdoor Unit). I had to have a Sontex because I run on 20% glycol, so the heat meter can to be calibrated accordingly.

I found the following in a recent test run:

  1. Temperatures into/out of the buffer tank fluctuate fairly rapidly as the various control parameters (HP controller outputs, TRV positions) take effect, so it’s not that easy to get steady state readings.
  2. My surface T/Cs gave outputs fairly consistent with the (nearby) heat meter T/C outputs (within ~1degC), and with the LWT and RWT reported temperatures (though there was a consistent ~1degC difference with LWT which I put down to either real temperature loss or T/C inaccuracy).
  3. Once the system was running reasonably steadily, the heat duty I calculated leaving the heat pump (from M.Cp.deltaT) was close to (with 5%) the heat duty I calculated passing the heat meter (and reported by it :slightly_smiling_face:).
    So in answer to your specific questions:
    A. Yes, by the sound of it.
    B. I’m sure that my secondary pump has been installed correctly!
    C. Typical temperature readings (once stable): T1 (ex HP) 48.9degC, T2 (to HP) 46.7degC, t1 (to emitters) 48.7degC, t2 (ex emitters) 42.8degC. My secondary flow is about 40% of the primary flow, so the deltaT (2.2degC primary, 5.9degC secondary) is fairly consistent.

So yes, the secondary flow (t1) should be reasonably close to your LWT (T1) but you should expect up to 1degC heat loss (LWT to buffer) on a cold day, depending on how much exposed piping you have and how well it is insulated.

(I realise that my secondary flow is not optimum - you should aim for primary and secondary buffer tank flows (thus deltaTs) to be about equal - but I’m still working on that particular case… :thinking:).

To summarise, I suspect that your perceived problems are less than your actual problems (at least with your buffer tank…).

Sarah

4

newplot(1)
newplot(1)934×450 44.5 KB

So, a picture is worth…
This is todays running. I think the legend is fairly self-explanatory but note that the background colour shows calling for heat, and the green-ish yellow area has a 2nd zone in play. “Pump Power” is the feedback signal from the secondary circulation pump (which is a UPM3 with nice PWM input for speed control and output of power consumption).

The space heating flow (SH-F) and heat pump return (HP-R) really are incredibly close! I think the periods when the pump is off are quite instructive. For example between ~12:30 and 13:00: a) the HP-F temperature (which is monitored very close to the cylinder) falls to the same value as the SH-R; b) the SP-F is highest. I interpret this as showing that the SH-F is at the top of the tank, while the HP-F is at the bottom. From about 12:30 it has established stable stratification.

During the period from 12:00 to 13:00, power consumption is only the HP quiescent 10W. Also, during this period, motorised valves are closed, so there should be negligable convective “turn-over” on the space heating side (I think this is evidenced by the gentle cooling curves for SH-F and SH-R during this period).

I cannot think of an alternative explanation than: the HP flow and return are connected to the wrong ports of the buffer.

I interpret the larger difference between HP-F and SH-F when both zones are an as being due to a larger volume of cool SH-R water entering the buffer and mixing with HP-F, while the HP flow rate remains roughly unchanged (or at least proportionally less changed)

1 Like
5

Thanks Sarah -
That T1-t1 difference is much more acceptable!

My aim was, as you note, to get balanced flows in primary and seconday by modulating the pump speed. Even if this comes down to some crude “rules” based on observation, as opposed to a fancy sensor-driven control loop, it would surely be an improvement. Getting the temperatures on each pipe was supposed to be the first step, and not the opening of a can of worms.