One of the biggest mistakes many builders make is to install a heating system fueled by propane or oil heat without considering an electric heat pump. In most cases that choice is costing the owners hundreds (or even thousands) of dollars a year in higher energy bills.
The graph at the right compares annual fuel costs for heat for a typical a new 2,000-square-foot Massachusetts home. A similar graph comparing the cost to operate different types of water heaters is shown below. The relative costs of the different fuel options will be similar for similar climates, but it’s important to note that natural gas is not available in many places.
Given these numbers, why are builders still installing propane and oil heating systems instead of electric heat pumps?
“I always thought electric heat was less efficient”
This is one common response I hear from builders when suggesting electric heat. This response is a product of history. Historically, heating with electricity meant resistance heaters. Resistance heat was common until the development and wide adoption of small, safe, and efficient fossil-fuel fired heating equipment.
Today, heating with electricity usually means using a heat pump, and today’s heat pumps are two to four times as efficient as a resistance heater. Heat pumps are completely different products from resistance heaters, and we shouldn’t be biased just because they use the same fuel source.
Is this a passing trend?
It’s impossible to predict exactly how energy prices will change in the next 10 to 30 years, but the U.S. Energy Information Administration (EIA) predicts that electricity prices will rise at an average annual rate of 0.4% between now and 2040, while oil prices will rise at 0.6% per year and propane prices will rise at 0.5% per year (excluding inflation).
This means that according to the EIA, heat pumps will only look more and more favorable in the coming years.
Some people don’t trust electric heat
Electric heat depends on the grid and the grid doesn’t always work. These are a few options for dealing with grid intermittency:
• Backup generator. This has the side benefit of running the lights and appliances during an outage, but it will cost at least a couple of thousand dollars.
• Backup heating system. The options here include a fuel-fired furnace, a boiler, or a wood-burning stove. Since it’s a backup system, efficiency can be sacrificed here to save money. A low-efficiency backup furnace system might cost about $1,000.
• Tough it out. If you’re confident in the grid and willing to throw on a couple of extra layers when there is a power outage, this is obviously the most cost-effective option. If there’s a risk of losing power for days at a time in the winter, however, you’ll want a backup system to avoid frozen pipes.
Heat pumps aren’t as efficient when it’s cold
When an air-source heat pump is working, it is actually pulling heat from the cooler outside air. As the outside air gets colder, the heat pump gets less efficient. For instance, today’s heat pumps will have a coefficient of performance (COP) of around 4 at 50°F and a COP of 1.5 at -10°F. This means that heat pumps will perform better in more temperate climates — although they don’t do that bad in cold climates.
The paper Climate Impacts on Heating Seasonal Performance Factor (HSPF) and Seasonal Energy Efficiency Ratio (SEER) for Air Source Heat Pumps (Fairey et al.) provides recommendations for accurately modeling the efficiency of air-source heat pumps in different climates. These recommendations are taken into account in the algorithm used to calculate the energy bill results shown in the graphs included above and below. High efficiency levels on temperate days more than make of up for low efficiencies during cold spells.
The issue of low-temperature efficiency is less important for ground-source heat pumps, because ground temperature is a lot more consistent than air temperature.
In addition to providing cost savings compared to propane and oil, heat pumps have some other important benefits:
• They are safe. Since there is no combustion in the house, there is less risk of air quality problems from carbon monoxide.
• They are convenient. No fuel tank or delivery is required, as is the case for propane and oil.
• They are clean. Fossil-fuel-fired equipment (especially oil-fired equipment) needs to be cleaned periodically. Cleaning is not necessary for electric equipment.
Air-source or ground-source?
Ground-source (or geothermal) heat pump costs vary significantly. In genera,l a ground-source heat pump will cost at least $10,000 more than an air-source heat pump but will be more efficient.
An energy modeling analysis and a financial analysis are especially valuable here, since this decision has a large financial impact.
Ducted or ductless?
The discussion at the beginning of this article assumed that a house has a forced-air heating system with ducted distribution. A heating seasonal performance factor (HSPF) of 8.5 was assumed for the air-source heat pump option, which is the equivalent of a mid- to low-efficiency unit — above the Federal minimum of 7.7 HSPF but well below high efficiency units like the Carrier Infinity 20, which boasts an HSPF of up to 13! The Goodman SSZ14 is an example of a commonly used unit; it has an HSPF of 8.5.
Ductless systems are more cost-effective and energy-efficient than ducted systems. They alleviate the need for costly, leaky ductwork, but they aren’t as aesthetically pleasing as ducted systems and it can be tricky to ensure conditioned air is properly distributed to all rooms of the house.
Relatively small homes that are well-insulated for their climate zone are the best candidates for going ductless. Builder Carter Scott’s homes are an excellent example of how to use ductless minisplits effectively in a relatively cold climate. His relatively small homes are superinsulated, and can be heated with a couple of minisplit units. For homes that aren’t as well-insulated, ductless heat pumps can also work great in auxiliary applications — like a bonus room, addition or apartment, where it’s desirable to have a separate, small heating and cooling system.
Heat pumps are the technology of the future
By allowing us to heat with clean renewal energy generated by photovoltaics, wind turbines, or a hydro plant, electric heating will be an important and necessary step in cutting greenhouse gas emissions.
Also, while fossil-fuel fired heaters have become about as efficient as they can possibly be, we are not very close to the theoretical limits of heat pump efficiency. So, heat pumps are only going to get better.
Nick Sisler is a co-founder and Engineer at Ekotrope, which provides software and energy consulting for builders. Ekotrope’s software is a RESNET accredited HERS and IECC Performance rating tool. Nick holds a bachelor’s degree in Mechanical Engineering from M.I.T.
Get building science and energy efficiency advice, plus special offers, in your inbox.
15 months later the math still works, eh? ;-)
Great minds think alike:
I'm skeptical of the model results for the heat pump water heater- it could be a matter of the input assumptions. The cost of the energy being pulled from the room air would have to be zero (or even negative- a presumed lowering of the cooling load?) for the annual hot water heating to be so heavily discounted with the heat pump water heater.
Here is a bar graph from the RMI report that Dana linked to. Note that (over the lifetime of the equipment) the cost of heat from a residential ground-source heat pump is more than the cost of heat from an air-source heat pump.
[Click on the image to enlarge it.]
Response to Dana Dorsett
Thanks for sharing the paper - I hadn't seen it before.
You are right to be skeptical of the heat pump water heater, the 2.4 EF assumes the water heater is in a warm room - which, in MA, means it would have to be in conditioned space, where it would be stealing conditioned air, increasing the heating space heating requirement. This was an oversight on my part. In MA a more realistic number would be 1.8 EF if the unit was kept in an unconditioned basement or garage. In those conditions, the expected annual bill is $274, rather than $204 shown in the chart. An updated chart is shown below.
Response to Nick Sisler
Thanks for the corrected graph. I have removed the original graph (Image #2) from the article and substituted your corrected graph.
Thanks for the explanation.
I figured it had to be embedded in the starting assumptions somewhere...
I'm also somewhat skeptical of the operating cost ranges of RMI's analysis, particularly regarding how similar the operating cost range of GSHP and ASHP are identical, which would imply they are comparing best-in-class ductless to middle-of-the road GSHP designs (and clearly not ducted ASHP in climates colder than US zone 4.)
But the rough gist of it is the same: Expensive fossil fuels like #2 oil and propane can't really compete with heat pumps on an operating cost basis any more. Given that the current levels of oil production requires an increasing share from oil sand or shale, which takes floor price of about $75/bbl for the producers to break even, heating oil really has no (affordable) future for home heating independently of the externality costs (that are not built into the purchase price.)
Also something to consider: Natural gas in burned in a combined cycle plant at ~50% net fuel-to-load efficiency heats a house at better than 100% source-fuel efficiency if leveraged by a heat pump delivering a seasonal average COP of 2 or better. About half of all annual kwh flowing on to the ISO-NE grid is from combined cycle gas, about a third is from legacy nukes, with the rest split roughly evenly between legacy ~30% efficiency coal (shrinking over time), hydro, and non-hydro renewables (increasing over time.) This makes heat pumps more source-fuel efficient, with a lower carbon footprint than condensing gas appliances in this region. As the grid greens up over time, the heat pump solutions green up with it.
But it's not endless bliss: Deployment of heat pumps at a rate faster than the combined increase in efficiency at the loads, expanded capacity of distributed renewable & storage could end up extending the service life and increase the operating capacity factors of the New England coal fleet, and deepen the demands on low-efficiency peakers. The more flexible the grid becomes, the easier it is to support a large increase in heat pump heating (and electric cars.) In the intermediate term deployment of micro-cogenerators are a natural complement to deployment of heat pumps for space heating, since the output of cogenerators naturally rises and falls with the local loading from heat pumps, reducing the generation requirements of large powerplants, while reducing substation congestion and transmission/distribution capacity problems.
These problems are soluble, but aren't exactly solved yet- regional energy policy needs to take into account all of the moving parts. Last winter there was some gas-grid capacity issues facing spacing-heating loads off against powerplant needs for the fuel, resulting in a gas price spike and at least one 15 hour open requisition in the day-ahead market during a cold snap, something of a wake-up call. Of course those drinking the frackwater see pipeline capacity expansion as the best solution (paid for by higher electric rates), but there are other voices out there, now that grid tied PV is widely seen as the soon-to-be-cheapest form of power generation.
Propane Marketing on the Ropes
I get pretty offended by stuff like this:
"Propane Appliances Save Money
Homebuilders who want to deliver the most energy-efficient, cost-effective, and safe homes for their customers choose to build their homes with propane"
@Dana, thanks for your thoughtful analysis and response! One point of clarification:
When you say cogenerators, I assume you're talking about cogeneration or Combined Heat and Power (CHP) plants that produce electricity and use waste heat for other purposes - typically space heating. Is that correct? These are great! MIT has one that provides electricity and (basically free) heat for most of the school. In that application - urban environment, single stakeholder - it works great. It doesn't work well in rural environments and is tricky in circumstances with a lot of stakeholders (like residential) - where it is tough to figure out how much each person should be paying. The DOE has a goal to generate 20% of electricity with CHP by 2030.
@Kevin, completely agree, the math just doesn't back those statements up.
Yes cogeneration (particularly MICRO-CHP) is great!
A few years ago the 1kw Honda CHP was being married to residential-sized condensing boilers/furnaces and sold as a system under the name "Freewatt". (The system was developed by a startup called Climate Energy, then sold to ECR International, who has since dropped the product line.) Honda is still manufacturing the basic cogenerator unit, but can't even meet the demand for the Japanese domestic market after the 44.2gigawatts of nukes went idle- don't expect to see them pushing to expand the market here any time soon.
Marathon Engine in WI has a ~4kw cogen called "EcoPower" with thermal output of reasonable size for single family homes and a 5kw Yanmar would be appropriate for larger homes, but for now both of those are primarily being sold into commercial space, not so much residential, and with very little local/regional support.
The 4-5kw systems are pretty expensive and have a tough financial rationale unless you have a big summertime thermal load (like a swimming pool.) But from a regional policy point of view these systems for space heating is good for grid reliability, and enables fossil-fuel defection by others in favor of heat pumps en masse without inducing a need for increased grid capacity. Together heat pumps & cogens are better than either alone.
Cogeneration plants aren't ever going to make any sense for single-family homes. However, they sometimes make sense for big multifamily projects or institutional buildings.
Never EVER? Tell that to the Europeans!
Given that power generators in New England vie for the same natural gas resources as home space heating, replacing a fraction of those gas-burners with distributed cogenerators would relieve the power-grid load AND the gas-grid load, by only marginally increasing the gas used for space heating, while feeding the grid during the space-heating peaks from heat pumps & air handlers, etc. It makes a HUGE amount of sense in New England, to avoid increasingly common repeats of this movie as more people use heat pumps.:
Europe has been on a cogenerator binge for decades, and in the past decade micro-CHP has become commonplace. There are probably getting onto a dozen 1000-5000W cogenerators being marketed for single-family home loads in northern Europe, eg:
To be sure typical thermal loads are lower and electricity prices higher than in the US, which does shift the financial picture a bit. But the capital expenditure on grid upgrades required to support major displacement of propane & oil space heating is substantial compared to subsidizing micro-CHP to make it viable.
NJ is now subsidizing micro-CHP (and other distributed power & grid-hardening measures) in the post-Sandy wake up. The 1kw Honda and a big buffer tank would be enough to handle the true loads of many NJ single family homes (and most condos), but SFAIK isn't available. That's really too bad, since it's a pretty cheap unit to build compared to some of the others, if higher maintenance (being a rankine cycle internal combustion unit) than the Stirling cycle cogenerators available in Europe & elsewhere.
Sales of backup generators have gone ballistic all over the northeast since Sandy- making them grid-connected but island-able cogenerators (like the versions of the Honda sold in Japan) would be a lot better expenditure of funds, for both the homeowner and everyone on the same local grid.
A quick web search turned up evidence that bankrupt startup Infinia's 3.5kw & 7,5kw Stirling technology cogens may soon reappear under new ownership:
It's not clear if Infinia will be marketed in the US (despite it's US origins) any time soon, since it's probably going to be an easier sell in Europe. The 3.5kw Infinia is sized adequately for most US homes (those without swimming pools anyway) without a lot of thermal buffering. It's peak thermal output is 48,000 BTU/hr- more than the design day heat load of most code-min 3000' houses (and many existing homes.)
edited to add:
I work with a guy who heats a large 1840s antique house with one of the early hydronic Freewatt systems. His annual gas use went DOWN (previously heated with an ~75% efficiency beastie-boiler) by over 30% and his net power use by more than half (on a ~20 cents/kwh utility). It paid for itself in about three years on the combined savings. The dollar value of the power offset was greater than the offset in gas use, and would have paid off in only 5 years on power savings alone, had the gas use stayed the same. As a system it is using slightly more gas than if he had replaced the cast iron beast with a modulating condensing boiler, but the "payoff" would have been over a decade without the Honda micro-CHP.
I also (and always) get the
I also (and always) get the question about "backup" heat, addressed in the article, but two of the options listed " a fuel-fired furnace, a boiler..." also typically run on electricity, so there is no difference between them and a heat pump. So the options are a generator, wood stove, or wait for inverter technology to catch up so you can use more than one outlet (in a grid tied PV system). That will probably happen soon.
About face on the grid?
A few years ago Ed Mazria and the folks at Architecture 2030 made what I thought was a really good argument: that architecture is the single biggest user of energy, and that the single worst kind of energy buildings use is electricity. As I remember, Mazria argued that while oil and propane are bad enough, coal is worse, and since the US grid is mostly coal fired and hopelessly inefficient, we need to work to reduce consumption of electricity, especially for low-grade applications like heating. (And along those lines there's that famous Amory Lovins quote about how using electricity for resistance heating is like cutting butter with a chainsaw.)
Fast forward a few years, and there is suddenly a big push to electrify. And not because the grid is any cleaner. I understand heat pump efficiencies help offset grid inefficiency (I think I've heard source-to-site numbers around 35% for grid efficiency in the NE?) but I am still a bit skeptical of this sudden about face.
Until the grid is cleaned up, electricity will remain high grade, dirty power. And in spite of the Obama administration's skewed argument for replacing aging coal powerplants with new gas fired ones, there is independent research to suggest that fracked natural gas is as bad as coal or worse when you account for fugitive emissions, birth defects, ground water pollution, radioactive waste disposal etc.
I wonder if anyone has looked at this in detail? Is a fracked-gas-fired air source heat pump really greener than the alternatives? I think I can see the benefit in a superinsulated home that needs only one or two for most of the year. But I've got clients in older houses who choose to put in multiple ASHPs instead of insulating. And what happens to electricity prices and supply when everyone piles on this heat pump bandwagon?
Seems like heat pumps risk being a trendy bandaid somehow--we are trying to solve a massive infrastructure problem by installing a new kind of appliance. In any case I think we should definitely be pushing hard for a cleaner grid if we are going to go down this road....
Response to Norm Farwell
With the possible exception of Ed Mazria -- I haven't spoken to Ed about this issue -- thoughtful green builders are moving in the direction of the all-electric home. Contrary to your implication, the grid is getting cleaner; coal plants are being retired, while wind and solar are ramping up. This transition is slower than most green builders would want, but it's happening.
And yes, many experts have made the calculations. Using ductless minisplits to heat your home reduces carbon emissions compared to the usual alternatives.
The graph below illustrates the declining importance of coal and the increasing importance of renewables in regards to U.S. electricity generation.
Real world ground source heat pump performance?
Several GBA articles have detailed higher than predicted costs and problems for ground source heat pumps, combined with lower than expected efficiencies, including testing protocols that mask some inefficiencies. This article seems to paint a fairly rosy picture of ground source heat pumps. I would appreciate a comment or two on how Martin, Dana, and others see the divide between modeled efficiency and installed efficiency for currently available ground source heat pump models.
Bob Irving writes:
"So the options are a generator, wood stove, or wait for inverter technology to catch up so you can use more than one outlet (in a grid tied PV system). That will probably happen soon."
It's not an inverter technology issue- islanding is relatively easy to design in, but a storage technology cost and regulatory environment issue. To manage peak startup surges requires at least a modest amount of local power storage, but many state regulations expressly disallow storage on the ratepayers side of the meter that isn't under the direct control of the utility, or onerous levels of permitting/review/metering/inspection to be able to install local storage. This issue was recently hashed out (and won, by local-storage PV proponents) in CA, which allowed solar installers (primarily Solar City) to move forward on coupling local storage onto localized PV with only very modest restrictions. See:
So "soon" is really "been there, done that"- it's a regulatory & storage cost problem, not a technical problem. The current darling of the small-scale storage market is lithium ion, which costs about $500/kwh, and for an all electric home to ride out a power outage of any length takes at least 5-10kwh of storage capacity- more if the space heating is heat pumps. There is a real possiblity that lithium ion storage will drop to $200/kwh by 2020, as the Tesla/Panasonic gigawatt battery factory gets built on schedule. But stationary storage doesn't need the high power-density of lithium ion- it can be bulky & heavy, but needs to be reliable. The liquid metal battery technology developed by Ambrii seems like a real contender for the distributed storage market, currently undergoing field testing at a wind farm in Hawaii. It's fairly low-tech, and has no moving parts, and it would't surprise me if it hit the $50/kwh price point once it is in full production. (Ground breaking on a serious sized production facility is coming this summer.)
norm farwell writes:
"I think I've heard source-to-site numbers around 35% for grid efficiency in the NE?"
That may have been true 15 years ago, but most of the new gas-fired grid sources on the ISO-NE and ISO-NY grids are combined cycle gas, running in excess of 50% thermal efficiency, which means coupled with a heat pump running a COP of 2.5 has significantly better source-fuel efficiency than a condensing gas furnace or boiler. The legacy coal units in the region run 25-30% thermal efficiency, but on the ISO-NE grid coal's share of the annual power going on the grid has shrunk to less than 10%, and that wiill continue to shrink with the scheduled closure of coal plants in MA over then next 2 years, and is being displaced by more combined cycle gas, offshore wind, onshore wind & distributed PV. Currently about 1/3 of the power going onto the ISO-NE grid is from legacy nukes, and that is unlikely to change dramatically (even with the imminently scheduled closure of Vermont Yankee at the end of the current fuel cycle.)
"And what happens to electricity prices and supply when everyone piles on this heat pump bandwagon?"
That is my lengthy argument about the policy efficacy of promoting micro-CHP in concert with the deployment of heat pumps. Since space heating and power generation in the region lean heavily on the natural gas resource, there is a peak-demand problem that can't be ignored. During peak heating events heat pumps are only running a COP of 1.5-2, and pulling a lot more grid power than on normal winter days. This became a problem in the ISO-NE grid during the January cold snap when gas-fired grid operators could not secure sufficient spot market gas (at any price!) to cover the peak load, and bidding in the day-ahead market for power (which is how most electricity is sold at the wholesale level) was open for fully FIFTEEN HOURS, (something that often closes in fifteen minutes.) Deploying distributed cogeneration en masse would alleviate this issue, since the micro CHPs would run at a maximum duty cycle during the same space heating peaks, delivering 90%+ net source-fuel efficiency while shaving the peak loads that need to be supplied by the large generators & distribution resources. It takes a lot of stress off the grid, and it's beneficial for the ratepayers to subsidize micro-CHP rather than paying for more low capacity-factor (operating duty-cycle) low efficiency peak power generators.
Distributed storage and smarter "demand response" (paid-for scheduled load-shedding by large power users during peak load periods) is cutting fairly deeply into the capacity factors of the existing 30-35% efficiency gas & oil fired peakers, during air conditioning peaks (when the stress on the gas grid is low, without the draw from space heating), and that trend is likely to accelerate as the grid gets smarter, and more responsive on the load end, and as the regulatory climate tries to catch up to the new technology/reality. NY has opened this can o' regulatory worms BIG TIME in just the past few weeks, which should enable a more rapid deployment of smarter and greener-cleaner use of the grid resources, economically accommodating far more competition from local grid-power & storage than is possible under the legacy regulatory structures. It's a big deal, but a bit dense for the non-grid-wonks to follow (even some o' them wonks are struggling with the implications.)
The ISO-NY & PJM (mid-atlantic and NE midwesternstates) grids still have quite a bit more legacy coal online than the ISO-NE grid, but there's little doubt that even with the amount of methane leakage on coal seam & shale gas the greenhouse footprint of combined cycle gas is far lower than coal- it's nearly 2x the efficiency and has only abou half the carbon per source fuel MMBTU, which means it's emitting only about 1/4 the carbon for the amount of power output. It's cleaner on other environmental concerns as well, though far from squeaky-clean. Leveraged against a heat pump's COP it's greener than burning the same gas in a condensing furnace.
You can watch the fuel mix change in quasi-real-time for the ISO-NE grid here:
The annual share for the ISO-NY grid for last year looks like this:
Coal is still king for PJM (which includes coal producing states like VA,WV, OH, PA, and a corner of KY.)
So it matters a bit WHERE you're hooking up your heat pump.
Response to Derek Roff
I stand by my previous writings on ground-source heat pumps. (See, for example, this article: Are Affordable Ground-Source Heat Pumps On the Horizon?)
The bar graph from Rocky Mountain Institute, which I reproduced in Comment #2, supports my position.
Energy-efficiency experts I trust, including Marc Rosenbaum, have shared many stories of screwed-up ground-source heat pump installations that take many hours of consultants' time to unravel and correct. There are no comparable ductless minisplit stories.
Moreover, I've interviewed two builders -- Carter Scott and Chuck Reiss -- both of whom have installed both ground-source heat pumps and ductless minisplits. In my article Living Without Electricity Bills, I wrote, "Going forward, Reiss will be changing his specifications for heating systems. 'We won’t be using ground-source heat pumps in the future,' he said. 'We’re going with air-to-air heat pumps — Mitsubishi ductless minisplits.'”
In my article called Just Two Minisplits Heat and Cool the Whole House, I wrote, "Before he discovered ductless minisplit units, [Carter] Scott built three homes with ground-source heat pumps (GSHPs). Now that he knows about minisplits, however, he has no intention of installing another GSHP."
you say thermal efficiency I say methane...
Thanks for the graph, Martin. And for the response Dana. I guess we see the same picture and read it differently. You see declining coal and increasing gas and renewables and a cleaner grid that plays well with heat pumps and chp.
But I look at the one line in Martin's graph that seems most out of sync with the others--that steep upward trend for natural gas consumption--and the alarm bells go off. Here's what concerns me:
1) The rate of increase in gas consumption (four fold since around 1990) and
2) The scale of it—the rise of gas approximately offsetting the drop in coal since 2005
3) Factor in credible independent research that says gas is worse than coal from a climate standpoint due to methane leaks--
In spite of the excitement, this looks like the revenge of the fossil fuel lobby to me. Meet the new energy, same as the old energy. Even the the recent EPA draft plan for reducing power plant emissions focuses only on C02 and completely ignores methane. Maybe that's just an egregious oversight, but I doubt it.
If the research by Robert Howarth and others about high rates of methane leakage in extraction and transport turns out to be accurate (and for what it's worth I see gas leaks everywhere as a home performance contractor, so it strikes me as plausible that industry might have the same issues on a much larger scale) then the improved thermal efficiency of new gas power plants won't be much help to us. And if so we "thoughtful green builders" might wish we had been more thoughtful.
Response to Martin/Derek (post #16)
The recently released -FH series Mitsubishi units will likely beat typical GSHP system efficiency in a New England climate if sized correctly. At mid-modulation these units run at a COP of about 3.5 at +17F (at max it's more like 2.5 @+17F). This is a huge boost over the state of the art ten years ago.
While there have been year-on-year improvements in GSHP efficiency, there has not been a corresponding year-on-year improvement in system designer/installer competence. Without the (not cheap or easy to do) system design optimization it's easy to cut the potential efficiency of this technology at the knees. Every system is a custom design, with many ways to screw it up. The cost of designing & drilling/digging has little chance of shrinking- it will always come at a premium.
By contrast a mini-split is a pre-engineered mass produced "system in a can", very well characterized, with no surprises, and (comparatively) low initial cost.
That's not to say you can't screw up with mini-splits- the more idiot-proof you make something the more creative the idiots become. But it doesn't take rocket science to size & install them correctly.
Given how far the ductless technology has come for cold-climate capacity & efficiency in the past decade I'm starting to think of small-scale residential GSHP in the past tense. The cost of doing GSHP right exceeds any marginal benefit. Even in implementations when that marginal benefit can be reaped, the energy use improvement is more cheaply offset by rooftop solar, even at 2014 pricing. So, even with the 30% income tax credit for GSHP it's not worth the design risk on new construction, provided the house can be designed & configured to be reasonably heated with mini-splits. For large scale new construction applications and for some older/larger single family home retrofits there may still be a financial (or practicality) case for GSHP, but for new small to mid sized single-family housing, not so much.
Howarth would differ on a couple of points
1) Long vs short view on emissions
Howarth's paper compares current average efficiencies for coal plants with current average efficiencies for natural gas plants. Dorsett looks at replacing the worst of the current coal plants with the best modern natural gas plants, and he takes a long view on emissions which diminishes the influence of methane pollution. In most things, I tend to take a long view, but not here. In this case, ignoring methane means ignoring the risk of shorter term climate tipping points. Howarth thinks "that's a risky proposition, since it would be disastrous if the climate system were to warm sufficiently in the coming few decades to hit a tipping point."
2) Not going directly to renewables is going to waste a lot of money
Scarce capital that could better be spent on renewable projects gets used up on dead-end fossil fuel infrastructure. Gas power plants are not short-term investments: with a typical 40 year life expectancy, they will tend to get used and maintained, and if not, rate-payers will be on the hook for the bail out. Current infrastructure will be extremely expensive to upgrade; Howarth points out that NYC alone has 3000 miles of century-old iron pipe that needs to be dug up or relined somehow. Critics also claim that drilling does enormous long-term economic and environmental damage to communities and regions where it happens.
3) Going directly to renewable is preferable
From a greenhouse gas viewpoint, going directly to renewables is far preferable as Jacobson et al. (2013) argue.
Dorsett tacitly assumes that natural gas infrastructure is necessary to support this transition to renewables. But he's not presented evidence to support that viewpoint, and there is good evidence to the contrary.
4) Price volatility
The price volatility for natural gas is very high, as the RMI piece notes. Tying electricity to gas will increase the volatility of the price of electricity. Unpredictable yoyo-ing energy prices destabilize the economy and are hard on businesses and families alike. Moving to renewables, on the other hand, should lead to much more stable prices.
Although I share the enthusiasm for heat pumps and I'm glad to hear Dorsett's optimism about the evolving grid, I think the idea that we need a natural gas build out to get off fossil fuel deserves serious scrutiny.
We agree more than you might think.
norm farwell writes:
"2) Not going directly to renewables is going to waste a lot of money. Scarce capital that could better be spent on renewable projects gets used up on dead-end fossil fuel infrastructure. Gas power plants are not short-term investments: with a typical 40 year life expectancy, they will tend to get used and maintained, and if not, rate-payers will be on the hook for the bail out."
Most cc gas plants on the ISO-NE grid are merchant power generators, not utility owned. The ratepayers are not on the hook, but the bond holders/shareholders can get stung pretty badly. It's pretty clear even to Barclay's analysts that bonds or other financial instruments for funding large fossil fired projects are increasingly risky due to the very serious prospect of competition from renewables. That's what I was referring to by my comment: "It remains to be seen if recently built cc gas plants will live out their design service life, even though they're absolutely necessary for keeping the lights (and heat pumps) running on the ISO-NE grid in the near term."
Vertically integrated utilities who own new-ish large scale plants are very likely to incur "stranded asset" costs, but it's up to the local regulatory bodies on how to deal with that. I'm very concerned about what will happen in Georgia, where it's a single large vertically integrated utility, currently building (and charging the rate payers a-proiri) large nukes that have little financial viability in the face of cheap PV + storage. Since they are a regulated monopoly they have largely been able to suppress/limit the amount of privately owned PV going on their grid, but if & when they go on line and the full cost gets built into the rate, grid defection becomes a realistic opportunity for the well-financed, leaving the high asset costs of the grid to be paid for by fewer (and poorer) ratepayers. This is a slow-moving disaster, but I'm not clear on how they will be able to turn the ship around in time.
norm farwell continues:
"Dorsett tacitly assumes that natural gas infrastructure is necessary to support this transition to renewables. But he's not presented evidence to support that viewpoint, and there is good evidence to the contrary."
Not AT ALL! I'm only assuming that the existing natural gas infrastructure is necessary for keeping the lights on TODAY. The transition is already under way and is accellerating.
I strongly believe that gas infrastructure in New England doesn't need to be built out further to make the transition, but I don't believe that transition can take place in 3 years either. (It's totally possible in 10 years and maybe even likely in 15.) Rather, I concur with NRG president David Crane's take on this:
"The purpose of having old coal plants, to be frank, is keeping the lights on for the next three, five, 10 years." "I'm not anti-utilities, I'm not anti-nuclear, I'm not anti-coal, I'm just anti-bullshit."
Rome wasn't built in a day (and if it were, I'd be suspect about the quality of the work, eh? :-) ) But I see the transition to a largely renewables ISO-NE grid to continue at an accelerating pace as PV & wind get cheaper just on the raw economics, even in the face of the fossil-fuel lobbies pressuring the New England state regulators & political bodies. The thin edge of the PV boom is currently only perceptible as negative load for the ISO-NE, but funny thing about exponential growth that doubles every two years- if it's 1% of the share now and 2% in 2016, in 10 years it'll be 16% of all grid kwh, and covering 100% of the air-conditioning peaks, and in 12 years it will be covering more of the fraction than the legacy nuke fleet. And that's just the currently hidden PV share- wind development isn't going to suddenly stop, even when PV becomes cheaper than wind. Wind is currently ahead of PV in the ISO-NE grid for annual kwh shipped, but my best-guess is that will not be the case in 2015, and maybe not in 2014.
Furthermore, I doubt the cc gas plant proposed to replace the retired coal plant in Salem MA makes economic sense. It's not clear that the new ownership (Footprint Power, a merchant power generation company based in NJ) will be able to get the financing for it, or that building it would even be in the interests of their shareholders. They may have gotten the Conservation Law Foundation withdraw their lawsuit to bar building it without further reducing scale & scope, but that's not to say it'll actually get built. Even if it does, the odds that it will operate for even the agreed upon reduced scale & period of time seem long. But that's now between Footprint Power, their shareholders & financial backers. Competition by renewables won't abate even if they build it, which is far from a certainty. The financial world has awakened to this new reality, and price lending accordingly. (A bit o' recent background on this particular situation: http://america.aljazeera.com/articles/2014/3/28/salem-power-plantsparkselectricdebate.html )
norm farwell continues:
"4) Price volatility The price volatility for natural gas is very high, as the RMI piece notes. Tying electricity to gas will increase the volatility of the price of electricity. "
That's ABS0LUTELY RIGHT! If gas hits a sustained $6/MMBTU well-head price (or higher) the financial case for even 20% capacity factor wind becomes viable, and can rapidly & economically hasten the expansion of wind power to displace that expensive gas fired power. And I believe that's exactly what WILL happen in New England. at $10/MMBTU large expansion of offshore wind becomes viable, but rapid expansion of PV + storage will likely undercut that even in the shorter term. I hope that mid scale offshore projects like Cape Wind get built- sooner the better, to offset nearby coal plant closings without increasing the capacity factors of gas or oil fired plants, but by 2025 like fossil-plants & nukes, the cost of NEW offshore wind has to come in line with the reality of lower cost distributed PV + grid storage. (Full disclosure: I'm related by marriage to Jim Gordon, the developer of Cape Wind, as if his being my brother-in-law's first cousin means anything. I've only met him a few times, at weddings etc. but was a fan of the project prior to knowing the connection.)
But in the meantime, micro-CHP makes financial and policy sense within the anticipated lifecycle of a CHP, and isn't any worse than (and is arguably better than) condensing gas furnaces, since it relives peak heating load grid strain without burning more natural gas (or oil.) In 15-25 years when the thing is finally toast, grid storage and a much greener grid will make replacing it with an (even higher efficiency )heat-pump will make sense. I see gas-fired micro-CHP as a bridge technology using the existing gas grid to reduce overall gas use, not as a 100 year solution to space heating & power. But there is little to suggest that within 10 years there will be sufficiently cheap & ubiquitous grid storage for all fossil-fired space heating replacements to jump directly to heat pumps, even if the current ISO-NE grid becomes all-renewables in that time frame, even if gas hits $20/MMBTU and even 10% capacity factor wind and even lousy shading factor rooftop PV becomes viable.
I'd love to be wrong on that though- we'll see. Things can turn around quicker than most people think. The grid load (and grid backup) from electric cars in 2025 could be substantial if lithium ion comes in under $150/kwh, which isn't out of the question. If you had told me 10 years ago that by 2014 in the state of Iowa something like 30% of all grid power would be coming from wind turbines I might have asked you what you'd been smoking. A lot can happen on grid evolution in 10 years.
Thanks for answering my question, and a typo
Thanks to Martin and Dana for responding to my question on the current wisdom/viability of ground source heat pumps for new residential construction. It's useful to have a mid-2014 confirmation of the information in earlier articles, that mini-splits have many advantages.
In comment 19, Dana says, "the installed capacity of PV is doubling roughly every 203 years." That looks like a typo to me, for "2-3 years". It might be worth an edit, since it is central to your point about the exponential growth of PV. (Or maybe you are taking the really, really long term view.)
Yep 2-3 years
... a much shorter time constant than 203 years, eh? :-)
The installed world capacity of PV (grid tied + off grid) is doubling at a sub-2 year time constant, comparable to that of grid tied in MA.
Distributed storage is poised for similar exponential growth, but there isn't sufficient data on how rapidly (and when) that deployment will occur.
Those looking to feed their inner policy wonk will want to download & read all 200+ pages of the full report from the links here:
But the rest may find the 13 page executive summary more useful:
response to norm farwell (#17)
Robert W. Howarth's article does not support the thesis that displacement of 25-35% coal with fracked-gas combined cycle gas is a net negative due to the increased methane quantities released, and the assertion than new-cc coal has a lower greenhouse gas footprint than cc gas has some real flaws. Yes, methane is a powerful, but short-lived greenhouse gas that needs to be better managed (and it can be), but the notion that the impact of the recently-sloppy methane management practices is worse in the year 2100 than generating that power with the lower efficiency of coal is not well founded.
The story is a bit different when comparing sloppy practices frack-gas burned at ~20-25% efficiency in an automotive internal combustion engine or a 30% thermal efficiency gas peaker compared to oil products. But methane releases can also be better managed & controlled than the recent-practices average, and Howarth's work has played no small part in awareness of that issue, bringing regulatory pressure to bear on the gas-biz.
But for the ISO-NE article makes two dubious assumptions. Quoting from the caption of figure 5:
"For the electricity production, average U.S. efficiencies of 41.8% for gas and 32.8% for coal are assumed."
While 41.8% may be a valid snapshot-in-time case for all natural gas fired power a national average basis, that is clearly not the case the combined cycle plants, OR the legacy coal units they are replacing in the ISO-NE grid. The legacy coal plants that are being retired are not new combined-cycle coal gasification units, they are typically operating at sub-30% thermal efficiency, whereas most new cc gas plants are averaging north of 50%. Furthermore, combined cycle gas is more flexible in output and can better track the 24/7 grid loads, never has to power-dump, and doesn't require must-run spinning reserves to track the load peaks, unlike the legacy coal plants. (Rigid slow ramping power plants have additional variable additional carbon footprint related to low efficiency must-run reserves for maintaining grid stability, a carbon footprint often overestimated for tracking the variable output of wind power.)
But the current uptick in methane footprint of of grid power related to natural gas may not ultimately matter, at least not for long. Both wind & PV will be both cheaper and more grid-resource efficient than combined cycle gas before 2025. Wind power already is cheaper than combined cycle gas at $4/MMBTU in many locations, but distributed grid tied PV will be cheaper than best-in-class wind in less than a decade. The grid storage party is just getting started, but it easily out-competes fossil-fired peakers on the raw economics of it. Only if the incumbent fossil powered generators win consistently in their regulatory lobbying will gas have more than a couple of decades of dominance (if that) in the power generation arena. The folks talking about natural gas powering the nation for the next 100 years have to be drinking the frack water, and ignoring just how viable both wind and solar are RIGHT NOW, let alone over the next couple of decades.
In Q1 2014 PV dominated the new-generating capacity in the US. While still a sub-1% fraction of the US total, the installed capacity of PV is doubling roughly every 2-3 years. This will probably have an inflection point in the exponent in 2017 when the 30% tax credit drops to 10%, but the cost of grid tied PV will be below grid-retail parity without subsidy before 2020. It really is ready for prime-time now, but will be noticeably cutting into the operating capacity factors of gas fired power generators in the ISO-NE grid by 2025.
In MA the recent PV capacity doubling has been faster than 2 years. Between the rise of PV and wind development some of the current plans for new fossil fired plants are at risk- they may not be able to get sufficient financing:
It remains to be seen if recently built cc gas plants will live out their design service life, even though they're absolutely necessary for keeping the lights (and heat pumps) running on the ISO-NE grid in the near term.
Between legacy nukes and increases in both hydro & non-hydro renewables, only about half the power on the ISO-NE grid is going to be from fossil sources in 2014. That's not even counting distributed PV on the other side of the meters, which only appears as negative load to the grid operator, and is not traded on the day-ahead or spot wholesale markets (though production credits from those installations are traded in the NEPOOL.). It'll be a different and even lower-carb mix in 5 years- count on it.
Log in or create an account to post a comment.Sign up Log in