Debate on Electric Heat Pumps

[Steve Macken]

Andy
- (firstly John, sorry to barge in.... !) the 2 key issues for my heat pump system were design of ground loop and design of UFH system.

As for the ground loop, the most economic installation is the slinky method, but of course this concentrates the collector in a relatively small area with an associated risk of the ground not being being able to replenish the heat lost to the collector quick enough from solar input (subject to ground conditions).  Out heat pump supplier (Nibe) advised against the slinky and reommended a 'snake-like' installation - more pricy as more excavation is required, but likely to more effective.

Secondly, the heat distribution should be designed for as low a flow temperature as possible, which for UFH is achieved by a large suface area of pipe (ie. large diameter and close spacing).  We designed for 35 deg. flow on the coldest days, whilst 27 deg seems adequate to maintain 20 deg. inside whilst it's freezing outside.  Of course this has an implication on material (more pipe) and labour cost, but this can be ofset by setting the pipes in the slab and not bothering with a screed.

Steve

[john cantor]

I have some info from the Powergen systems. These are based on Calorex heat pumps, and were designed for social housing projects. They are either 3.5kW or 5 kW (output).  usually one borehole per house.
I will try to find a link to this info.
One site, of 10 houses, has been monitored for 3 years, the second site, of 15 houses, for 2years.
An accumulated total of 600,000 hours of service has been logged,
The average Seasonal Performance Factors for the heat pumps were calculated as between 4.0 and 4.25 when supplying underfloor heating, and between 2.25 and 2.5 when delivering high temperature water to the primary coil of the hot water cylinder.  The DHW figure could be better, these heat up to around 60C. (no back-up)
We also have monitored the water source heat pump at CAT. total figures were looking at just under 4.2. this is with a source of around 8C on average , supplying large radiatior at the reception. In theory the COP should be more like 4.8 and I dont know why it is lower. Note this heat meter is not callibrated.
In the case of cat, the COP is higher than a normal GSHP because the water entrers at 8C on average. many GSHP systems have water from ground at anything from 2 to 5 in the winter..or worse. Many systems I fit operate higher due to more source pipe. My one at home works mostly around +7C due to ground water. .
Extra pipe does not cost much, but its hard to make a living by taking that approach.
I still maintain that batch heating mains water from 12 up to 50, rather than cycling it at 45-50 should result in DHW COPs of 3.5ish. But I really should get my finger out, fix my heat meter and run some tests. This approach requires much bigger storage cylinder...more cost.

[David OLIVIER]

John

These COPs of 2.4 for hot water and 4.1 for space heating are useful.

A very well-insulated and draughtproofed house may need say 2500 kWh/yr of heat delivered to the hot water tank and 2000 kWh/yr of space heating delivered to the rooms.

On that basis, the expected weighted average COP = 3.1. This is marginally above the lower limit of 3.0 required in the present version of the Energy Standards.

A COP of 3.1 for a heat pump driven from a coal-fired power station with an overall efficiency of 0.3 from primary energy to delivered electricity isn't very good; it's 0.93 which is less good than a good condensing boiler plus radiators (Gordon Taylor an AECB member, has measured the efficiency of his boiler at 96% for the last 7 years). It only looks good if the electricity to drive the heat pump comes from a gas power station, with a much higher overall efficiency of say 0.40 from fuel to delivered electricity. However, I doubt this represents the actual situation. Due to high gas prices, new generation investments to meet the UK's rising electricity demand (both peak capacity and annual energy sales are rising) mostly take the form of coal-fired plants (wind capacity is fairly low and is rising more slowly than peak demand is rising). When you add the extra cost of underfloor heat, versus radiators operating at say 55/35 degC, the extra cost of underfloor insulation to reduce the downward heat losses from the heated slab, and the extra cost of installation, I very much doubt the heat pump can compete with other CO2-saving investments.

Some COPs are undoubtedly being quoted by producers which attempt to exclude hot water (which other heating systems provide) and to exclude resistance heating backup (which lowers the overall COP). All figures should as far as possible be quoted on a like for like basis. Exceptions to this should be clearly justified.

David.

[john cantor]

David,
Its always interesting to read your deductions.   

I am surprised by your quoted Gordon Taylors efficiency measure of 96%. I had thought that steady-state laboratory tests gave just over 90%, and in the same way that it seems that the COPs of heat pumps are less in practice than manufacturers data. I have believed that the same applies to boilers.  My experience of heat meters is that they can vary by over 10% in either direction, and they don't capture intermittent readings very well due to lags in sensor response. 

Most of my comparisons are not with gas, but with oil boilers , since that seems many peoples second choice in remote areas,

could you remind us of your thoughts of actual UK co2/kwh figures for mains electricity.  (i.e. Defra say 0.43kg/kwn.)   I recall that you though UK was about 0.5, and Germany was actually 0.55!

I reckon gas is .205kg co2/kwh at best (93%?). with Electricity at 0.5, this gives a break-even COP of about 2.5, which obviously needs to be exceeded by a significant amount to make a heat pump worthwhile.
Oil at 90% is about 0.3 co2/kwh , so breakeven COP 1.7.
LPG being about 10% better.
Lots of houses around here use oil.

How we actually assess the pros and cons has puzzled me for many years.. and it continues to do so.

[David OLIVIER]

John

Gordon Taylor measured his 96% on a system with a modulating boiler, weather compensation and continuous heating; i.e., normal good practice controls across the Channel. The heat emitters are standard panel radiators, operating at a higher temperature drop than normal and designed to give a low return temp. No underfloor heating is present or needed; it's a 1965 detached house. I think he's posted some of the monitoring data on his site energypolicy.co.uk.

We need a costed UK supply surve of CO2-saving and energy-saving measures, with any extra features needed by, say, a heat pump clearly set out, so that we can see how different options compare to each other. If you do this calculation for a heat pump replacing a condensing oil boiler, the cost in £ per tonne CO2 displaced seems to be relatively high compared to other CO2-saving investments; e.g., improved windows with lower U-values. You can't spend the same money twice (spending it on a heat pump could hypothetically take the budget which had been allocated to upgrading the windows from a U-value of 1.2 to 0., so you need to spend it wisely.

DEFRA have quoted a figure of 0.527 kg/kWh since about 2006. See http://www.aecb.net/forum/index.php?topic=1064.0. These are the annual average emissions and matters get more complicated when you want  to know winter emissions or the implications of measures which increase electricity demand, such as heat pumps.

David.

[Mark Siddall]

The attached graph is a replication of something I saw in a paper by Bob Lowe. I feel that the graph works well to help conceptualise the issues arising from the COP and Carbon emissions debate.

I've tweaked it to show the CO2 emissions to reflect with some recent data used in the CSH addendum.

Mark

[David OLIVIER]

Mark

New gas supply in the UK now takes the form of LNG from Qatar, where c. 20% of the gas is consumed for liquefaction / re-evaporation.

New UK elec supply (vital, given that consumption is rising 10x faster than renewable supply) is from coal, not gas. Winter elec;  i.e., the season when well-insulated buildings consume 80-90% of their space heating energy, is worse than the national average.

Would like to allow for these issues.

Bob's graph also excludes CHP plant (can be equiv to a heat pump with a COP of 10-14) and IMO the higher COP of 4 is usually impracticable, given the Carnot limit and the fact that small devices don't get very close to this.

David.

[Mark Siddall]

David,
My understanding of the graph is that it, broadly speaking, can accomodate an increase (or decrease) in Carbon Intensity of Elec (kg CO2/kWh). All that happens is that the "UK Carbon Intensity" will shift to the left (or right).

That the Carbon Intensity of Elec is indeed a concern and not one that should be ignored (on the contrary in fact) but what the graph does enable is an apprecaition of the fact that as the grid is decarbonised so technologies such as heat pumps become more appropriate as a solution.

So whilst I agree that what we really need is a more up to date/accurate data on the current Carbon Intensity of Elec, so far i have not found anything that goes beyond the data presented.....

Cheers,
Mark

P.S. Incidentally I note that one of the CEPHEUS reports actually report at seasonal COP for MVHR of 13 (If it were not for the fact that the UK climate is more mild than Germany my addition to Bob Lowe's graph could, in this context, be considered to be conservative).