Geothermal Operating Costs vs. Gas Furnace

Discussion in 'Maintenance and Troubleshooting' started by mts18geo, Apr 23, 2013.

  1. mts18geo

    mts18geo New Member

    I’m looking for some perspective on the electricity consumption and costs of a new geothermal system that has been operating for the last six months. The good news is that the system kept our home warm for through the winter, but the electricity costs are significantly higher than what we were previously paying to heat our home with natural gas.

    Here’s the situation
    * Home is in Western NY with approx 3600 sqft on two finished floors. Basement is unfinished.
    * Last winter (’11-’12), the home was heated by two 95% efficient gas furnaces with variable speed air handlers
    * 6ton Advanced Geothermal Technologies DX system was installed last Fall, and was operational for this winter
    * In January, I realized that our gas & electric bills were much higher than last winter and saw that our household electricity usage had increased 2500-3000 kWh per month compared to the prior year. (Gas usage was much lower than last winter, and nearly flat from the summer – which is what we expected.)
    * We then installed a separate meter to track the heatpump (and desuperheater) electricity usage so we can isolate from other household consumption.
    * When I do the math of heating cost and adjust for HDDs to get a cost per HDD, I find that the costs of the heating with the geothermal system are ~60% higher than they were with natural gas. (This is based on the current natural gas costs. If I use last year’s gas price which was higher, the geothermal costs are still 45% higher than gas when I adjust for HDDs.)

    Needless to say, this is concerning because we made a very large investment in a system with the expectation that it would save us 50%+ on energy bills each year, and right now it looks like we’re paying more each month and would be better off just running the gas furnaces.

    My take is that one of three things is happening:
    1) Something about the system isn’t operating correctly, and it’s leading to much higher than normal energy consumption to run the system. This was my first instinct, and also hope, since that would imply that it could be fixed and we’d be back on the path of saving energy/money. However, the system seems to be cycling on/off at reasonable intervals – when it runs, it’s typically for a relatively short period. Not sure what else could be wrong.

    2) Something about the particular system design/size/configuration isn’t right for the situation – e.g., is system/compressor is larger than needed, and therefore consumes extra energy? Not sure what system/design choices could lead to big increases in energy consumption without additional symptom of system running excessively.

    3) Or, perhaps the high efficiency furnaces were already providing heat at very low cost, and a geothermal system shouldn’t be expected to perform at even lower costs – in other words, investing in geothermal was a bad idea in our case where the baseline heating costs were already low.

    I can provide more details of the system and measurement/calculations if necessary, but I’m thinking folks here may have a good general sense of what kind of energy/cost savings would be expected when moving from natural gas to geothermal.

    This experience has been disheartening – I’d been aware of geothermal systems for a while and was excited to have the opportunity to install at my home. I was looking forward to have a positive story to share with other and maybe win more converts. However, at this point I’m starting to feel that this was a big waste of money, but hopefully that’s wrong and someone can help me figure out what’s going on.

    Thanks in advance for any thoughts or advice for next steps in diagnosing or fixing.
  2. I think you have correctly identified the possibilities, it would be helpful to know some more about your system and its operation.

    * Does your system have an electric resistance heater that might be coming on to supplement energy from the DX loop?

    * What kind of electric meter are you using?

    * Can you provide kWh over a time frame for this past season and the natural gas consumption over a period from last year?

    * It would also be helpful to know your local fuel rates, I would assume something like $12/Mcf for gas (winter) and $0.16/kWh for electric, are those correct?
  3. AMI Contracting

    AMI Contracting A nice Van Morrison song Industry Professional Forum Leader

    Yes gas and electric costs are a must. Heat load calcs would help too.
  4. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    Where are you located and who installed the system.

    DX systems don't keep up the promise, but also someone telling you that you would save 50% of nat gas with high efficient boilers is setting expectations too high, at $8.50/MCF and 12 cents/KWH, which are the going rates right now in Western New York.

    You are welcome to check performance data from system installed in the Western New York area.

    See link below

    Buffalo GeoThermal Heating
  5. mts18geo

    mts18geo New Member

    Here are the answers...

    1. No, there isn't an electric resistance heater. The supplemental heat source is the gas furnaces, which I don't believe were called much (if at all) this winter. Our gas usage was pretty flat compared to the summer.

    2. The electric meter is Centron -- similar to the type of meter we have outside our house. It doesn't track hourly or more granular energy usage, but I've been reading it every few days since mid-march to get a sense of the usage pattern.

    3. Last Winter: Gas usage from 12/27/11 - 2/27/12 was 386 therms (6.2 therms per day)
    Last Winter: Electricity usage from 12/27/11 - 2/27/12 was 1592 kWh (26 kWh per day)
    This winter: Gas usage from 12/29/12 - 2/27/13 was 97 therms (1.6 therms per day -- slightly higher than we used during the summer -- we have gas water heater and stove)
    This Winter: Electricity usage from 12/29/12 - 2/27/13 was 6952 kWh (116 kWh per day ... 90 more kWh per day than last year)
    Note that these meter readings above are for the full house. In mid-march we installed the separate meter to track just the geothermal system, and for the first week or so when it remained cold the measurements were 88 kWh/day -- consistent with the +90 we saw compared to last year for the house. More recently it has averaged ~36 kWh per day as the weather has warmed.

    4. Current energy costs are approx. $.80/therm and 9.6 cents per kWh. (This is the full variable costs of the energy -- delivery + supply.)

    And to answer another question below, we are in Rochester NY.
  6. engineer

    engineer Well-Known Member Industry Professional Forum Leader

    So, you essentially traded consuming 4.6 therms of natural gas per day (gross energy content 460,000 Btu, cost $3.68 / day) for 90 kwh / day (gross energy content 307,170 Btu, cost $8.64 / day)

    That's a raw deal, and I'd be pissed.

    I'm in Florida, and we had a really mild winter, but news reports from places where water frozen solid often falls from the sky and persists on the ground for months at a time suggest that it was wicked cold up theyuh this yeayuh, so paht of your experience may stem from substantially increased heating load this year over last. You could compare degree day data between this year and last at Rochester Airport to get a handle on that factor. National weather service should have that climate data online for your reference. I routinely use NWS JAX (Jacksonville, Fla.) data when similar questions arise in my area.

    Weather aside, consider the cost of delivering a given amount of heat to your home: A million Btus is a common reference quantity - In your case getting that million from 80 cent therms of natural gas burned at 90% efficiency would cost $8.89. A ship-shape geo system using 9.6 cent kWh operating at an average coefficient of performance of 3.0 would deliver a million Btu for $9.38.


    However, I'm not convinced that switching to geo is a complete fail. I don't expect 80 cent therms to persist given the efforts underway to retrofit roughly 12 US ports to export liquified natural gas to markets where it costs 4x what it now does here.

    All that said, it does seem that your geo fell far short of delivering anywhere near a COP of 3.0, and that needs to be thoroughly investigated. I would seek a local contractor (possibly a certified air balancer with the tools to measure air flow and power consumption) to measure actual COP while there are still some cool nights left to work with. That would consist of measuring real time power consumption (accounting for power factor, not just Amps X Volts, of all system components) and air flow and temperature rise supplied to that air flow.

    If I was there I'd also seek refrigerant conditions, pressures, temperatures, superheat and subcooling to try to get a handle on what is happening on that side of the system. My guess is that something is awry on the refrigerant side evidenced by higher than design compression ratio, longer runtimes and lower than design heat transfer to airside, but that is strictly conjecture, free-and-worth-every-penny-paid distant internet advice.

    Good luck! - keep us advised.
  7. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    This year's heating degree days were 30% higher than last year's in Buffalo, NY, about 50 miles away from you.
  8. AMI Contracting

    AMI Contracting A nice Van Morrison song Industry Professional Forum Leader

    That moves the numbers closer but yikes cutting the gas bill in half was an ambitious goal. I'd take Curt's suggestion and get someone to have a hard look at the operation side. I would try to get the manufacturer involved as well.
    If you want some insight into DX look here:
  9. Rosco

    Rosco New Member

    I'm also having a DX system put in (on the other side of the planet, so my perspective may be upside down to yours). I also had a keen desire to confirm the real performance of my system but didn't want to wait for my next bill or season to get an answer. So the simple but effective method I used just looks at the time it takes to heat (or cool) the water in the hydronics buffer tank and calculates the energy transferred to (or from) the tank. I then compare that directly to the electric energy consumed in the process.

    My method:
    My system's hydronics buffer tank has a volume of 400 litres plus a few more litres in the hydronic circuits through the house.
    With the exchanger off I run some hydronics circuits for about 30 minutes just to mix the water in the buffer tank to get a reasonably even temperature throughout. Turn them off while the measurement is made.
    I turn the system on and wait for it to start up then take three readings (buffer tank temperature, time of day, kWh on the power meter)
    I then wait for the tank temperature to reach just shy of the set temperature where the system will turn off and take the same readings again.
    Now I have start and end temperature, time of day, and input kWh.
    Now I calculate the energy required to change the temperature of the buffer tank water volume. All this requires is the specific heat capacity (Cp) of water, which is 4.186 (using SI units). I.e. to raise 1 ltr (1 kg) of water by 1 deg celcius (or kelvin) requires 4.186 kiloJoules of energy.

    Cooling mode illustration:

    Start readings were:
    Time: 8:42
    Temp: 33.10 deg C
    kWh: 33,277.42
    Stop readings were:
    Time: 9:22
    Temp: 10.00 deg C
    kWh: 33,279.62

    From this I get:
    input energy (from meter reading) = 33,279.62 - 33,277,42 = 2.20 kWh
    output energy (from temp change, Cp, volume) = 23.10 * 400 * 4.186 / 3600 = 10.74 kWh
    Output to input ratio = 4.88

    I can get values for average instantaneous energy consumption by dividing by the time:
    input power (kWh divided by run time in hours) = 2.20 / 0.67 = 3.30 kW
    ouput power = 10.74 / 0.67 = 16.12 kW
    This might be useful, e.g. as a sanity check input power should be close to what is documented for the system.

    A few comments:
    1. This measurement can be done anytime to track system performance under different conditions or over time.
    2. It is independent of how the house or the interior hydronics is performing, so can quickly isolate where a problem may be occurring. E.g. a very low ratio could indicate a problem with the refrigerant gas, or that the ground around loop field is thermally saturated/depleted.
    3. Determining the volume of water in the entire hydronics side of the system may be important for an accurate result, e.g. unless you have check valves in these circuits there will be some back flow through them while the system heats (cools) the buffer tank. Any such backflow will affect the result slightly but should hopefully be negligible.
    4. The ratio of output to input, which should be a little higher than the COP rating of the system since it doesn't measure end-to-end performance, should normally vary slightly depending on conditions, e.g. if the system has been running for a while or just starting up after period of extended inactivity, etc. For example, I would expect the ratio to be higher if the loop field has been idle and lower if the loop field is in more of a steady state condition, where the system has been running for a while.
    5. It may also be useful to run this test with one or more or different hydronic circuits running. I haven't tried this.
    6. If you are also using your Geo system to heat your DHWS, make sure you run this test outside of any scheduled DHWS heating cycles. If using the cooling cycle to run this test it is probably better to ensure your DHWS is already topped up and you don't use any hot water during the test, unless you want to specifically test that aspect of your system.
    I would welcome any comments on this, particularly if there are errors.
    Last edited: Dec 16, 2015
  10. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    Entering and leaving temperature on both the ground loop and the load have significant impact on the performance. As a rule of thumb, a 4.5 C higher temperature on the ground has 15% higher performance as a result (in heating mode).
    In the same matter your COP will be different when your buffer tank is 33.1 C versus 10 C. Strinkingly different. That itself can make a 50% difference. Add to this the warmer ground temperature after the heat heats up the ground. especially DX system swing very quickly in terms of ground temperature. At start up they look sensational, but after the ground temperature swings quickly after a couple hours, they usually don't look so good anymore.
    Last edited: Dec 17, 2015
  11. Rosco

    Rosco New Member

    I hope I didn't sound like I was evangelizing DX over water loops. This method would. I hope, be useful for either.

    But thank you for the numbers and insights because I'm hoping to be able run this test under different conditions and over time, and comparing those results will give me a better feel for when things aren't quite right.
  12. Rosco

    Rosco New Member

    Last weekend here (Summer - Sat 32 deg C, Sun 35 deg C) provided an opportunity to run the system in cooling mode continuously both days except for a few hours in the early morning and so in light of docjenser's comments I took the following readings at the start and end of one cycle on the Sunday where the system was re-chilling the buffer tank. During this cycle all zones were turned off. Only one zone was running all weekend and the call on the system all weekend was about 1/3 of it's capacity.

    Start readings were:
    Time: 14:43
    Temp: 14.2 deg C
    kWh: 33,646.60
    Stop readings were:
    Time: 14:58
    Temp: 9.30 deg C
    kWh: 33,647.32

    From this I get:
    input energy (from meter reading) = 0.72 kWh
    output energy (from temp change, Cp, volume) = 4.9 * 400 * 4.186 / 3600 = 2.28 kWh
    Output to input ratio = 3.17

    This is still not entirely accurate as I'm not sure how to treat the start and end temps. The set temp is 10 deg and the margin is 4 deg, so the system starts up when the tank temp reaches 14 deg and stops at 10 deg. There is a lag in the system so the temp continues to rise to 14.2 at the start and continues to drop down to 9.2 deg after it stops. If I use the trigger temperatures I get a power ratio of 2.58 rather than 3.17. The true value is probably closer to 3.

    I'm an interested geo customer, not an installer, so this next may be entirely off the mark. Words of experience are very welcome.

    The system has a published heating COP of 4 and I know that cooling COPs are always lower, which I'm reasoning is because:
    • The internal energy transfer COP for the system is assumed to be the same regardless of heating or cooling mode, which is 4.
    • In heating mode the heat of compression adds to the heating of the tank, so 4kW of heat into the tank requires 1kW from mains and 3kW of heat from the ground. That is, an effective heating COP of 4.
    • In cooling mode the heat of compression adds to the heating of the ground, so 4kW of heat into the ground requires 1kW from mains and 3kW of heat from the tank. That is, an effective cooling COP of 3.
    If that is true then 1kW of cooling requires 1.33kW into the ground, whereas 1kW of heating requires 0.75kW from the ground. That's a seasonal ground load asymmetry of 1.33/0.75 = 1.77. That is, per kW, cooling places a much higher demand on the loop field than does heating. If loops fail in summer then this may be the reason.
  13. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    Yes, compressor heat needs to be rejected in cooling mode, and adds to the capacity in heating mode.

    Yes, cooling places a much higher demand on loop fields, but mainly because you have the phase change releasing a large amount of energy in form of heat when water converts to ice. This is the main reason loops stabilize at 32F (0 degrees Celsius).
  14. Rosco

    Rosco New Member

    That doesn't sound right and i can't find a way for it to make sense.

    I can see where the ground would stabilize at zero C in heating mode, i.e. while removing heat from the ground, because the ground water turns to ice which supplies heat to the system without dropping the ground temperature further. I assume you are talking about a DX loop here.

    Conversely, if there happened to be ice in the ground in summer (unlikely) a system in cooling mode transferring heat to the ground, converting that ice to water would absorb heat without the ground temperature rising.
  15. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    You must read up on the physics.
    Indeed, when ice turns to water it consumes a large amount of energy in form of heat, that is why snowmelt takes up a huge amount of energy, even more when turning from water to gas, which is called evaporation.

    I had copied you a link to a physics page explaining the whole thing.
  16. Rosco

    Rosco New Member

    The bit I don't get is: where in a geo exchange system running in cooling mode does water become ice in or around the loop field? Under what conditions might this happen?
  17. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    Ice does not form in cooling mode in the loop, since heat is rejected, only heating mode, where I was referring to.

    Again, much more strain on the loop in cooling mode, since the compressor heat gets added, and the loop heats up quickly. However, in heating mode, the loop gets not strained so much, since a portion of the energy comes from the compressor, and the phase change ensures that it does not drop below 0 degrees Celsius very much, unless it is grossly undersized.
  18. engineer

    engineer Well-Known Member Industry Professional Forum Leader

    I guess that answers my Q on another thread asking why lopsided cooling loads are more difficult to cope with.

Share This Page