Hello @iantelescope thanks for these, could you give a bit more detail on your monitoring setup? you mentioned using an Arduino above, is that using a CT sensor or reading from a meter? How are you doing the temperature sensing?
Would it be difficult for you to share the above data as a CSV file? ideally if you have 10s data that would be great. It would then be easier for us to zoom in on the periods with the oscillations. Ideally if you have flow and return temperature separately.
It would be easier if you simply pasted the image into a post rather than using PDF.
Hi Trystan and Brian,
The Following represents my , hopefully, unambiguous , description of my current measurements.
I will buy , or build a better measurement structure when I have more money!
Power Measurement
I measure the Power using a Current Transformer ,CT, clamped across the Heat Pump Mains input.
The CT output Voltage is then fed into an Arduino.
The Arduino “program” performs no Filtering or Windowing on the measured voltage.
The Arduino then calculates the RMS Energy in Joules over a 10 second measurement period.
The RMS Energy over 10 seconds is then stored as a measure of power when multiplied by a Calibration Constant to give RMS Power.
The RMS Power is then put into an accumulator as a measure of RMS Energy.
Both Power and Energy are stored on an File on an SD card,.
Power and energy data
1619.csv (43.7 KB)
The Data consists of the following space separated data fields
Year Month Day Hour Minute Second Power Energy.
Temperature Measurements
Four K type Thermocouple Clamps are attached to
1 The Heat Pump Output
2 The Heat Pump Return
3 The Radiator , or Heat Exchanger Output.
4 The Radiator, or Heat Exchanger Return.
I have stored the Temperatures using a four input K type Temperature Logger.
Temperature Data
01061939 .csv (51.9 KB)
The Temperature data is stored on an SD card in the following Format
AT 2023-01-06 19:39:34 1m 29.3 31.6 30.2 30.3 C
where
2023-01-06 is here the year - Month - Day
1m is the Temperature Sampling Period
29.3 is the Heat Pump Output Temperature
31.6 is the Heat Pump Return Temperature
30.2 is the Heat Exchanger or Radiator Output Water Temperature
30.3 is the Heat Exchanger or Radiator Return Water Temperature.
The Measured Temperatures are here the external Temperatures of the pipes!!
The Temperatures shown here are NOT, therefore, the Water Temperatures within the pipes themselves !
File Formats
I will in future show the Output Files as both PDF and Jpeg.
For 9th Jan to 10Jan the files are now
and
power.pdf (9.4 KB)
Energy
energy.pdf (16.2 KB)
Absolute Temperature
abs_temp_HP.pdf (9.9 KB)
Delta T HP water Temperature
del_flow_temp.pdf (10.5 KB)
Heat Exchanger ( Radiator )
del_rad.pdf (11.0 KB)
Hysteresis?
Could the Oscillations or Cycling be the result of the Hysteresis setting of the Water Law Thermostat?
The Hysteresis setting appears to be set by selecting from
1: Field option 2091 Option 0 . Neither Water Law Thermostat nor Room Thermostat Switch used.
2: Field Option 2091 Option 1. Room Thermostat only used.
3: Field Option 2091 Option 2. Room Thermostat AND Water Law Thermostat used with Hysteresis = 0.
4: Field Option 2091 Option 3. Room Thermostat AND Water Law Thermostat used with Hysteresis = 1.
5: Field Option 2091 Option 3. Room Thermostat AND Water Law Thermostat used with Hysteresis = 2.
All of these options work on my Heat Pump,
I will experiment with changing the Water Law Thermostat Hysteresis, 2091 options
ian
Ian, thanks - this is a bit more usable.
My interpretation of your data is:
Of course I may be wrong but I saw similar patterns on Glyn’s public dashboard, especially when his flow temperatures were very low.
What water law rules were recommended for your system?
Hi Christian,
Perhaps my neurotic contemplation of My Heat Pump performance would be allayed with Power , Energy and Temperature Graphs from other users?
The Oscillations are clearly real in that they consume lots of Energy, but are the Oscillations common with other users?
Are these Oscillations the result of Hysteresis?
Many years ago , a Greek engineer, perplexed by my use of “Hysteresis” translated Hysteresis as “madness” .
Perhaps my Heat Pump is “mad”?
Neither my “installer” nor "Dalliam Samsung " nor “Telford” have ever advised or recommended any system settings for my "unique " system.
ian
Have a look at the Samsung 5kW one, click on All Time then click on the column for 7th October, I think you will see similar performance.
Hi Christian,
Again, a revelation. Many thanks.
The Graph shown is excellent …and triggered thought.
My “unique” system has a 50 litre “Volumising Tank” fitted across the Primary ports of a Plate Heat Exchanger.
The Heat pump, when supplying heat to the Radiators has a Water Volume of 56 Litres.
The Heat when supplying the Hot Water tank has a Water Volume of 6 Litres.
Therefore, the Time constant , or Cycle time , of the Heat Pump, is nine times longer when supplying the Radiators than the Cycle Time when supplying Hot Water.
The Hot Water supply Cycle time will be nine times shorter than the Space Heating Cycle Time.
With the Volumising tank fitted across the Primary of the Heat exchanger
Energy Consumed = Flow T x Sp_HT x Volume = 35 x 4.3 x 50 = 7525 Joules = 7.525 KJ.
With the Volumising tank fitted across Heat Pump
Energy Consumed = Flow T x Sp_HT x Volume = 50 x 4.3 x 50 = 10750 Joules = 10.75 KJ.
Therefore the Volumising tank was fitted across the Primary of the Heat Exchanger.
The Hot Water will therefore oscillate with a 9 times higher frequency than the Radiator Feed.
QED
ian
Thanks Ian and @christian!
Is the heat exchanger used instead of a coil in the hot water tank? or is it between the heat pump and the radiators? any chance you could draw a diagram of how it’s connected up?
Technically yes, but if considering the thermal mass of the operating mode [while heating the water] you will need to consider the water tank + heat exchanger interaction with the 6 litre volume of pipe. [I am assuming when heating the hot water, the heat exchanger is used, and when not it is direct to the radiators]
For reference here is my flow temp data (pump is on the flow), the spikes are when heating the hot water tank (I have a different setup to yours of course as I have re-used my existing 300l copper vented tank - with some modifications to make it a bit like a mixergy tank)
my main observation is that my flow gets back to normal very quickly, and does not jump around as much as yours, though there are many differences between how I am measuring and how you are measuring. In anycase you may find this useful.
Hi Trystan and William.
Again many thanks.
Please see the attached drawing , of a by now, very complex Heat Pump.
Samsung-inchinnan.pdf (20.1 KB)
William , again you are technically correct, in that Thermal load of the Hot Water Tank should be added to the 6 litre pipe volume. My Comment was only intended to show the differences between the Radiator Cycle Time Constant and the Hot Water Cycling Time Constant.
My Water Cycling oscillations only occur when heating Hot Water.
My Power Cycling oscillations only occur during De-icing.
The Heat Exchanger was fitted in order “to isolate the Heat Pump from Variations in Flow rate and pressure on the radiator circuit”.
The Volumiser tank was fitted “to reduce the cycling experienced when the Heat Pump was delivering at it’s minimum output , 25% , of the Maximum Output Power” during Sept/Oct 2022.
A later statement from Dalliam Samsung required the Volumiser tank for frost damage prevention.
Both of my manufacturers , Samsung and Telford, required the Volumiser tank to be fitted.
The Calculations of the Volumiser volume are shown in the MCS “Domestic Heat Pumps , A best practice Guide” page 35 under “Buffer Vessels”.
Using the " MCS Best Practice " I calculated the required Volumiser tank Capacity as 35 Litres.
Samsung Dalliam were going to "install the Volumising tank and issue a Warranty certificate , but at an further large cost of circa £5000.
A 50L Volumising tank was installed in late October by my “installer”.
The MCS "best Practice Manual " page 35 has issued a Warning.
WARNING from MCS.
“Buffer Vessels can perform important functions if incorporated correctly, However they can also increase system heat loss.”
The Hot water tank can climb to 50 C where the Energy stored would be
Energy loss in Volumiser if fitted across HP out and HP return
Energy ( Joules ) = 50 x 4.3 x 50 = 10.75 KJ
Energy loss in Volumiser if fitted across the Heat Exchanger Primary ports
Energy loss ( Joules ) = 35 X 4.3 x 50 = 7.725 KJ.
The loss being least when the Volumiser was fitted across the Heat Exchanger.
ian
HI William,
Again , many thanks for the data from your Heat Pump.
The Differences between your Temperature Measurements and my own , are vividly contrasted.
My Temperature Measurements are the Temperatures as measured on the outside of the pipe, while yours are presumably taken inside the water flow itself.
My Temperature measurements are also taken using K type Sensors , which are both extremely fast and extremely sensitive to Electrical Noise.
I will have to get your , in pipe sensors asap!
I have been pondering the question " at what temperature should a Backup Gas heater be activated ?".
My Electricity costs 33.763 p per Kwh .
My Gas would cost 10.276 p per Kwh.
The Electricity to Gas Ratio is therefore E/G = 33.763/10.276 = 3.28.
The Temperature at which the Heat Pump COP = E/G would set the point at which the Gas Backup should be started.
Looking at the COP to Temperature charts
Note that my PLR , (Power Loss Ratio ) = Operating Power / Maximum Power = 2Kw/5Kw = 0.4.
I would conclude that the Gas boiler should start when the temperature goes below -5 C.
As an aside, I have noted the YouTube video describing the 201 - 203 * Field settings containing the comment
" The default settings, 201-203 are Expensive and should be modified to -3 C and +13 to +17 C depending upon the House insulation.
The user " should, however, experiment to find a suitable balance between Room Temperature and Economy".
The Video announcer then says
**Users in Scotland should use -5 C and +13 to +17 C , but experiment **
Again many thanks for your help
ian
Hi William,
I am now content that I understand the cause of both the Hot Water cycling and the Energy Pulsing.
I can explain , and justify, everything, except the Water Temperatures, with the Water Temperatures requiring a proper, immersion sensor array.
I can explain All , but can do nothing!
I have asked the three other “installers” who originally quoted for the job for their opinions.
All three have asked for varying sums of money to “repair, and or replace”, the Heat Pump, with the largest , taking £96 for their “advice”. Another wants to , not only replace the Heat Pump, but also the Pipes and the Radiators! While the final contender has launched a personal attack on me , blaming me for my “stupidity in selecting” such a parcel of rogues.
Were it not for the intervention of my MSP I would be in serious financial trouble as well as cold!
I should have , like yourself, built the Heat Pump myself!
I will , however, build a proper Temperature measuring array for the Water Temperatures on the Heat Pump.
ian
HI Trystan,William and Christian,
Why is Energy measured in Kwh ?
Energy in the Gas industry is measured in MJ/m^3 ,Mega-joules per Cubic meter.
Power is measured in Kw , Kill-Watts in all industries.
The printed characters Kw is very similar to the printed characters for Kwh.
The Confusion would be stopped if All industries, including the Renewable industry, measured Energy, ,unambiguously, in Mega-Joules , MJ
MJ looks completely different to Kwh.
Kwh looks too similar to Kw.
Conversion
1 kwh = 1000 J/sec over 1 hour = 1000 X 60 X 60 Joules. = 3.6 MJ.
Why NOT use MJ for Energy!
Use Kw FOR Power !
ian
How is Lord Kelvin involved? K is the symbol for degrees Kelvin. And what about James Watt - isn’t he deserving to be honoured by a capital letter?
My guess is, as a Chartered Electrical Engineer, is it’s down to history, and kW & kWh are far more convenient for us, as they don’t involve multiplying and dividing by a pesky 3.6 or 3600 every time we measure a current and a voltage. OK, a joule is one watt-second, but an inconveniently small unit for most purposes.
Hi Robert,
The Joule
Such a heat unit, if found acceptable, might with great propriety, I think, be called the Joule, after the man who has done so much to develop the dynamical theory of heat… Siemens.
I was hoping you would respond!
I appreciate the history of James Watt , William Thomson ( Lord Kelvin ) and James Joule.
However, the use of the Watt , the Kelvin, and the Joule should be open for debate.
I am not proposing the Joule as a measurement of Energy …
I am proposing the Mega- Joule , MJ
The “pesky” 3600 multiplier would be replaced with a multiplier of 3.6.
The MJ would emphasise the environmental cost of using too much energy.
A bill in Mega-Joules would be sure to frighten the punters!
ian
Thanks, It would be interesting to monitor the flow and return temperatures either side of the heat exchanger and before and after the volumiser. There is a fair bit of debate as to the benefits or drawbacks of installing low loss headers, buffers or heat exchangers, you might find this page and accompanied heat geek video interesting Low Loss Headers: The Complete Guide For Efficient Design, heat pump manufacturers usually specify hydraulic separation as a precaution as they dont always have control over what kind of heating emitter configuration is connected to the heat pump and so need to make sure that there is adequate flow at all times… regardless of how the user has set their TRV’s etc. It might be worth collecting more data on your system before passing too much judgement. It is possible to get systems with hydrolytic separation to work well, though my experience with the use of heat exchangers for this is limited…
I personally have a heat pump without one of these sources of hydraulic separation and instead ensure that there is always adequate flow and system volume by keeping all radiators in the house fully open at all times. Here’s the diagram of my system:
It has worked well for me now for 3 years and counting, providing a SCOP of 4.08, here’s the monitoring data dashboard Emoncms - app view dashboard. Glyn’s samsung installation is an identical configuration.
The COP required by the heat pump to break even is not quite that high, as the gas boiler is not 100% efficient. This study found that the mean efficiency for combi boilers in the trial was 82% https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/180950/In-situ_monitoring_of_condensing_boilers_final_report.pdf
That drops the break even COP required by the heat pump to 33.763÷(10.276÷0.82) = 2.7
That boiler efficiency does not as far as Im aware include the boiler electricity consumption, central heating pump and controls which are all usually included in the heat pump SCOP values.
I wish I could! 
That’d keep my electrical bill to a minimum.
