musingsheader image
3 Helpful?

Solar Thermal Is Really, Really Dead

A new analysis shows how much cheaper it is to heat water with PV modules than with solar thermal collectors

Posted on Dec 26 2014 by Martin Holladay

Back in early 2012, in an article called “Solar Thermal Is Dead,” I announced that “it’s now cheaper to heat water with a photovoltaic(PV) Generation of electricity directly from sunlight. A photovoltaic cell has no moving parts; electrons are energized by sunlight and result in current flow. array than solar thermal collectors.”

Now that almost three years have passed, it’s worth revisiting the topic. In the years since that article was written, the cost to install a photovoltaic (PVPhotovoltaics. Generation of electricity directly from sunlight. A photovoltaic (PV) cell has no moving parts; electrons are energized by sunlight and result in current flow.) system has dropped significantly. Moreover, I’ve come across monitoring data that allow for a more accurate estimate of the amount of electricity needed to heat water with electric resistance elements or a heat pumpHeating and cooling system in which specialized refrigerant fluid in a sealed system is alternately evaporated and condensed, changing its state from liquid to vapor by altering its pressure; this phase change allows heat to be transferred into or out of the house. See air-source heat pump and ground-source heat pump..

First, I’ll present my assumptions.

How much does a solar thermal system cost?

In my earlier article, I estimated that a residential solar thermal system with two 4' by 8' collectors and a solar storage tank with a capacity in the range of 80 to 120 gallons costs between $8,000 to $10,000 to install. I stand by that estimate.

Of course, some contractors can beat this price, while others will charge significantly more. (In a recent comment posted on GBA, an Ohio-based solar contractor named Daniel Young estimated that the solar thermal system I describe would cost $16,250.) For the purposes of the comparisons made in this article, I’ll assume that the installed cost of a residential solar thermal system is $9,000.

How much does a PV system cost?

My calculations are based on a PV system cost of $3.74/watt. The figure comes from a the “Solar Market Insight Report 2014 Q2” published by the Solar Energy Industries Association.

Some GBA readers have received quotes of $3.50/watt for a PV system, while others are still paying $4.00/watt or more. One thing’s for sure: prices for PV are still dropping.

The price comparisons made in this article do not include any incentives, rebates, or tax credits.

How can I determine the annual electricity production of a proposed PV system?

The easiest way to figure out how many kWh will be produced each year by a PV system is to use a free online calculator called PVwatts. The calculator allows users to change a number of parameters, including geographical location.

How much does a water heater cost?

Installation costs vary from region to region. This article assumes that the installed cost for an electric-resistance water heater is $1,200, and the installed cost of a heat-pump water heater is $3,000.

Of course, your local costs may be higher or lower than these figures.

How many gallons of hot water per day does the average family use?

According to a Canadian study, the average Canadian family uses 44 gallons of hot water per day. The Canadian researchers’ findings mirror those of several U.S. researchers; there is growing evidence from monitoring studies that the assumption used in the DOE’s Energy Factor test for water heaters — namely, that an American family uses 64 gallons of hot water per day — is unjustifiably high.

This article assumes that the average North American family uses 44 gallons of domestic hot water per day.

In homes with a solar thermal system, what percentage of the home’s domestic hot water needs are met by the solar equipment?

One of the best studies on the “solar fraction” question was performed in 2006 by researchers from Steven Winter Associates. The researchers monitored two residential solar thermal systems for a year, one in Wisconsin and one in Massachusetts. Each house had two solar collectors. According to the researchers’ report, Cost, Design and Performance of Solar Hot Water in Cold Climate Homes, the solar fractions of these two carefully monitored systems were 63% and 61%, respectively.

I’m going to assume that a two-collector solar thermal system supplies 63% (on an annual basis) of a family’s domestic hot water needs. (The energy used to heat the remaining 37% of a family’s hot-water needs is provided by a backup water heater — for example, an electric resistance water heater).

Some solar thermal enthusiasts may argue that there are locations in the U.S. where the solar fraction for a solar thermal system is likely to be higher than 63%. They're right. However, it's important to remember that in locations with lots of sunshine, the annual output of a PV system will also be higher than it would be in a location like Wisconsin or Massachusetts.

How much electricity is needed to make domestic hot water?

The electrical energy use assumptions in this article are based on data provided by Marc Rosenbaum, who has monitored the energy use of several Massachusetts families for years.

According to Rosenbaum’s monitoring data, a typical electric resistance water heater uses 0.21 kWh/gallon of hot water (3,373 kWh/year to make 44 gallons of hot water per day), while a typical heat-pump water heater uses 0.07 kWh/gallon of hot water (1,124 kWh/year to make 44 gallons of hot water per day).

Comparing three systems

My latest approach to comparing the cost of solar equipment used to make domestic hot water starts with the assumption that the typical solar fraction of a cold-climate solar thermal system is 63%.

Keeping that solar fraction in mind, I have calculated the cost of equipment for three scenarios:

  • House A has a solar thermal system (two rooftop collectors and a solar storage tank in the 80 to 120 gallon range) and an electric-resistance water heater for backup.
  • House B has an electric-resistance water heater and a PV system sized to provide enough electricity on an annual basis to meet 63% of the family’s hot water needs.
  • House C has a heat-pump water heater and a PV system sized to provide enough electricity on an annual basis to meet 63% of the family’s hot water needs.

The table below compares three homes in Boston, each of which uses 44 gallons of domestic hot water per day.

According to this analysis, the PV plus electric-resistance approach is about 25% cheaper than the solar thermal route, and the PV plus heat-pump approach is about 50% cheaper than the solar thermal route.

Everyone's numbers are going to be different

What if you are a solar thermal buff who thinks that my assumptions are unfair to solar thermal? Well, let’s change a few numbers. We’ll assume that a solar thermal system costs only $6,000 to install and that the solar fraction is 75%.

I think that the assumptions made in the above table are unrealistic, since it's hard to find a contractor willing to install a good two-collector solar thermal system for $6,000, and because in a location with a solar fraction of 75% a PV system is likely to produce more electricity than this table shows. But according to this analysis, the PV plus heat-pump approach is still about 23% cheaper than the solar thermal route.


Using the information in this article, GBA readers can perform their own calculations. For some readers, the cost of a solar thermal system will be higher than either of the above analyses. For others, the cost will be lower than my lowest assumption. But most results will be similar to the results shown above.

One more point to consider: if your numbers result in a tie — if your calculations show that solar thermal equipment and PV equipment cost exactly the same — remember that the maintenance costs for a solar thermal system will be higher than the maintenance costs for a PV system.

Some people love their solar thermal systems

Solar thermal technology attracts a loyal group of fans. Solar hot water systems have been around for more than a century, and for renewable energy enthusiasts of my generation, these systems evoke fond associations. I have two solar thermal collectors on my roof, and I enjoy listening to the quiet hum of my system's small pump when the sun is shining.

Solar thermal buffs make the argument that solar hot water systems are simple, elegant, affordable, and delightful. I sympathize with their emotional attachment to these systems, and I wish them all the luck in the world. However, most people I know who have installed solar thermal equipment have a tale or two involving maintenance headaches. (If you are a hobbyist, these technical glitches are sometimes exciting to solve. If you are an average homeowner, however, these glitches are just ordinary headaches.) Owners of PV systems are much less likely to have as many stories of maintenance problems.

Martin Holladay’s previous blog: “Martin’s 2014 Christmas Parody.”

Click here to follow Martin Holladay on Twitter.

Tags: , , ,

Mar 12, 2015 4:21 AM ET

Edited Mar 12, 2015 5:01 AM ET.

Response to Graham Irwin
by Martin Holladay

I'm happy to concede one of your points -- that "On a cost basis, the relative benefit [of a PV system] is dependent on the vagaries of the utility billing scheme in question."

But cost is almost always a factor in almost all construction specifications. After all, the reason that most of us don't install gold roofing (the type of roofing installed on the State House in Montpelier, Vermont) has to do with cost, not performance. Your observation that cost is a factor is not particularly original, and does not disqualify my arguments.

These costs are subject to change, of course. Not only are PV modules getting cheaper, but some electric utilities want to reduce the credit they now provide to homeowners who sell them PV-generated electricity. It's up to each homeowner to determine where they stand in this shifting landscape, and whether an investment in PV makes sense.

In some U.S. states, including Wisconsin and Arizona, net metering contracts are PV-hostile. In other states, including California and Massachusetts, net metering contracts are PV-friendly. It's hard to predict which approach will prevail in the end. If this uncertainty is enough to sour you on PV, then you shouldn't invest in a PV system.

Your statement that "the US grid is, for the most part, a coal and natural-gas-fired generator that never turns off and delivers electricity at about 33% efficiency" muddies the facts. The percentage of electricity that is generated by coal varies widely from region to region. Where I live, in Vermont, the contribution from coal-fired power plants is 0%. In fact, the carbon intensity of the grid is falling, due in part to the growth of PV, but mainly due to the increase in wind-generated electricity and the substitution of natural gas for coal. This reduction in carbon intensity should be celebrated; moreover, we all have a chance to take part in this grid transformation by installing a PV system.

Right now, PV systems supply less than 1% of America's electricity, which is a far cry from the 20% that makes utility executives (and you, evidently) quake in their boots. When a PV system feeds electricity into the grid, all of that electricity displaces electricity that would otherwise have to be generated by conventional power plants or other means.

The electricity grid is imperfect and polluting, but most of us use it. I imagine that you use it, too. Although my own home is off-grid, my lifestyle (like the lifestyle of almost all Americans) depends on the grid in many ways. Since we depend on this grid, pointing out its imperfections is useful, but only up to a point. Encouraging the installation of PV systems, in hopes that a much higher percentage of our nation's electricity will be produced by PV in the future, is a good thing, with more benefits than most analysts realize or admit.

You wrote, "I do agree with you that solar thermal systems are often too expensive in this country." This fact (which we agree on), coupled with the falling cost of PV, will soon make the debate on these pages irrelevant. Americans aren't particularly eager to spend $8,000 (or more) for a solar hot water system, but PV installations are booming. Economics will trump any debate over plumbing vs. electricity; fortunately for the planet, the PV boom is part of a necessary transition to a low-carbon future.

Mar 12, 2015 12:41 PM ET

by Graham Irwin

A couple Gedanken (“thought experiments”) while I'm ruminating on the topic:

1) Suppose a solar thermal system is designed for a house where the yearly output of the collectors matches the estimated domestic hot water demand. The solar tank contains a backup heating element with the thermostat set to 120ºF. The system has a sensor to prevent overheating that directs the solar energy to a pool, or a spa, or the domestic hot water for an in-law unit, etc. This sensor is also set to 120ºF. Assume that all of the thermal energy collected by the system is put to use. What percentage of the domestic hot water for the house is heated with solar?

2) Imagine two identical homes, side by side, identical occupants, identical schedules and usage. Both homes use electricity for water heating, identical water heaters. One home has a PV system installed, the other does not. After morning showers, dishwashing, etc., we watch the electrical meters on each house. Is there any difference between the rate of change of the readings of the meters on the two houses as the water heaters draw electricity from the grid to replace the hot water? In the late morning to early afternoon (no one is home) the house with the PV system sees the meter spin backwards, the other house does not. Evening comes, and people come back from work and school and use more hot water, then the water heaters draw electricity from the grid to replace the hot water. Once again, is there any difference between the rate of change of the readings of the two meters? The house with the PV system has made an environmentally beneficial contribution to the grid at mid-day, but how can one say that the PV system heated their water?

Mar 12, 2015 1:43 PM ET

Edited Mar 12, 2015 2:11 PM ET.

Response to Graham Irwin
by Martin Holladay

Concerning question #1: the solar fraction of a solar hot water system depends on the climate and the size of the system. Most two-collector systems for single-family homes end up with solar fractions in the 60% to 70% range. Increasing the solar fraction from (for example) 65% to 75% or 80% is possible but expensive -- and as you try to reach 90% or 100%, you end up needing lots of collectors and a very large storage tank in all but the sunniest climates.

In your scenario, a system that produces the same amount of hot water on an annual basis as the family uses will obviously not result in a 100% solar fraction. When there are three weeks of cloudy weather in November, the solar fraction will drop to 0% or 5% for those weeks. When there are three weeks of brilliant sunny weather in June, the system will make more hot water than the family can use, and the system will dump lots of unneeded hot water into the dump loop.

Your second question -- about "the rate of change of the readings on two electrical meters" -- confuses me. What do you mean by "the rate of change"? On a partly sunny day that starts out cloudy, a 2 kW PV system will go from producing 20 watts at 11:00 a.m. if it is cloudy, and, in 5 seconds, when the cloud blows away, the system can suddenly start producing 2,100 watts. That is a very fast rate of change -- from 20 watts to 2,100 watts in 5 seconds. Speedy Gonzales.

Mar 19, 2015 8:47 AM ET

remove existing system???
by Darin Smith

We moved into our house 2 1/2 years ago (northern climate). It had an existing solar thermal system just hooked up to the water heater (electric backup). The previous homeowner installed this himself and is no longer in the area. I can't find anyone in the area to work on it or tell me if it's working. I now need to replace the roof in a few months, so it would be the ideal time to get this hideous thing off my rooftop. So, knowing that I didn't really pay anything out of pocket for the I just keep it? Or get rid of it now to avoid expensive maintenance down the road?

Mar 19, 2015 9:13 AM ET

Response to Darin Smith
by Martin Holladay

Ideally, you would find an experienced solar contractor to evaluate the system and tell you whether it's working. If that's impossible, I can't think of a single reason why you would want to keep it.

You may be able to sell the used collectors to a hobbyist or do-it-yourselfer.

Mar 19, 2015 12:35 PM ET

PV panels are great, but they aren't "magical."
by Graham Irwin

Please explain to me how the PV panels in your example are "heating water." It seems to me that in most situations, hot water is drawn and replaced with cold water in the morning before there's any significant output from the PV system, and the water is immediately reheated with (mostly) coal and natural gas burned at the power plant at ~30% efficiency. When the occupants come home from work and school in the evening, the process repeats. Yes, the PV output in the middle of the day is environmentally beneficial, but it's probably powering commercial buildings, correct?

Mar 19, 2015 1:10 PM ET

Edited Mar 20, 2015 5:16 AM ET.

Response to Graham Irwin
by Martin Holladay

If you're thinking of an off-grid house, you're right, of course. You would need a battery to store electricity generated between 9:00 a.m. and 4:00 p.m. if you needed to use the electricity in the evening or early morning.

But that's not how things work in a grid-connected house. You can't actually attach a label to an electron at 10:00 a.m. and figure out where it goes. Electricity isn't like that. Every building connected to the grid draws electricity from the grid as it is needed, and all of the PV arrays and coal plants that are connected to the grid feed electricity into the grid.

If you wanted to connect your heat-pump water heater or electric-resistance water heater to a timer, so that it only operated between the hours of 9:00 a.m. and 4:00 p.m., you could certainly do that. If you did, you'd probably need a bigger water tank. This hypothetical system would behave somewhat like a solar thermal system -- although of course it would be mechanically simpler and much more reliable.

If PV-generated power ever reaches the point where it is supplying 18% or 20% of the electricity needs of the U.S., utilities may need to develop new strategies to handle the PV-generated electricity. But we're not at that point yet -- and if we do reach that point, it will be good news for the planet.

Mar 19, 2015 4:29 PM ET

Magical thinking
by Graham Irwin


I think there's a fairly strong correlation between grid energy used and environmental impact, i.e. grid electricity consumed = fuel burned @ power plant = environmental impact. There is no "big battery" (except hydro-electric in some situations) in the grid so what's really happening is that the atmosphere is being used as "storage" in your proposal - carbon is emitted during water heating, at a later time, the PV system is reducing potential emissions from another use in (likely) another location. Even this "storage" analogy is a bit of a stretch as the PV doesn't sequester carbon, it relies on synchronized potential pollution to offset.

Why is this anything other than a semantic difference? Because it can lead to illogical conclusions. For example, if the "net zero" water heating approach has no environmental impact, it would be impossible to reduce the environmental impact. In fact, if the identical PV system were retained, and the water heater exchanged for something that ran on gas (or even coal) at an efficiency greater than the grid (~30%) x the COP of the heat pump water heater, there would be less environmental impact, no? This doesn't take away from the benefit of the PV system, it merely delineates the two activities rather than conflating them.

Beyond that, if one counts avoided emissions at another time as equivalent to emissions from the activity itself, why is PV the holy grail? Could one not, for example, convert one's garage into a home office, get rid of the commuter car and telecommute, and then claim "credit" for that activity as an offset against one's water heating? How about forgoing red meat? Both of these carbon reduction approaches are extremely cost-effective. If someone published an article stating that "veggie burgers are a cheaper way to heat water than PV" I think it would likely be met with ridicule, yet it follows the same logic.

Mar 19, 2015 4:56 PM ET

Edited Mar 19, 2015 5:04 PM ET.

Response to Graham Irwin
by Martin Holladay

If you can find a way to live off the grid, be my guest. I'm doing it.

If you are using grid electricity, welcome to the club. We all want a cleaner grid. One way to get there is to increase the percentage of PV-generated electricity that is contributed to the grid. Right now, that percentage is extremely low in the U.S.

I never claimed that PV is magic -- although it's close. (The first time I hooked up my PV modules to a 12-volt grain grinder, and watched the sun turn wheat into flour on a sunny day, it certainly seemed like magic to me.) Nor did I claim that PV is the holy grail, or that the grid includes a big battery.

I will continue to report, however, that every kWh of PV-produced electricity that is fed into the grid displaces electricity generated by other means -- which in the U.S. includes electricity generated by coal, natural gas, and nuclear power plants.

Of course we need to reduce our energy use as much as possible, and live environmentally responsible lives. If you can reduce your carbon footprint by working in a home office and eating rice and beans, I support you 100%.

People who live in a grid-connected house in a location that allows homeowners to sign up for net metering can reduce or eliminate their electricity bill when they install a PV system. This investment in a PV system yields better energy savings, and a better financial return, than an investment in a solar thermal system.

These facts don't prevent you from installing a solar thermal system. You can have one if you want. I do.

And if you remain seriously hung up about whether or not your PV production is simultaneous with your electricity use, I strongly suggest that you put your electric water heater on a timer so that it only operates from 9:00 a.m. to 3:00 pm. or 4:00 p.m. Then you can stop bugging me about simultaneity.

Mar 19, 2015 8:38 PM ET

by Graham Irwin

Put the heat pump water heater on a controller so it raises the set point when there's output from the PV panels, add double the area of PV than comparable solar thermal panels (assuming an average COP of 2) and you may have an interesting proposition there. If you want to use electric resistance, make it 4x the area.

If you stop making questionable claims that "bug" me, I will refrain from commenting here and use my time more wisely - deal?

Mar 20, 2015 4:40 AM ET

Response to Graham Irwin
by Martin Holladay

I'm glad that you found my suggestion interesting. I'll take that as a positive note, so that we can end our conversation on a note of agreement.

If you take my interesting suggestion, there will be many benefits. In addition to reduced maintenance costs, a huge benefit would be the fact that once the aquastat on the water heater is satisfied (because the tank is up to temperature), the PV array would begin spinning your electric meter backwards. That's a trick that no solar thermal collector is capable of.

Mar 20, 2015 7:37 AM ET

existing system
by Darin Smith

Thanks Martin for your quick response yesterday on my existing system that the previous homeowner installed. Is there an easy (safe) way of turning this off for a month so I can compare electric bills? Short of that, I don't know how else to really determine if this is saving me money or not.

Mar 20, 2015 9:24 AM ET

Response to Darin Smith
by Martin Holladay

I don't know whether your system circulates a glycol solution, or whether you have a drainback system, so it's hard to know if it can be safely shut down.

The best way for you to shut down the system would be to get up on the roof and to cover the collectors with a heavy tarp, secured so that it can't be blown away. That's safe.

If your system has glycol, you don't want to close any valves in the collector loop, because the glycol can overheat and be ruined. Nor do you want to cut electrical power to the circulator, for the same reason.

Mar 21, 2015 10:05 PM ET

Great article, stimulating comments, BUT
by Janis Bell

In the hottest part of California, in the Coachella Valley desert, solar thermal for DHW is indeed dead. I called every company in the area, and all respond about maintenance hassles due to the intense heat on rooftops in the summer. On the other hand, at a recent Home Show, there were 4-5 vendors selling solar coil collectors for pools -- a simple solar thermal system that tends not to break in the heat.

Another deterrent to solar thermal, from the only company I found in a 100 mile radius, is their requirement for a 1500 gallon storage tank to supply enough hot water for those pipes and DHW use in the winter. Where does one find room for a 1500 gallon tank? Not in anything I'm going to build.

We are already getting backlash from utility companies. I am planning a residence in the high desert where heating costs tend to equal cooling costs, propane is $4 a gallon, yet the local electric cooperative can't accept any more net metering because the AZ company that sells them electric requires they purchase a certain quantity to continue to get their low price.

I'm still hoping to find a solution that allows me to harness as much solar power as possible for all my needs. The solar gurus discourage battery storage due to the short life of batteries in the heat and say their clients living off grid pay $5000 every 2-3 years to replace them. The dominance of PV net metering sure makes it hard these days for someone in a sunny climate to make use of all that free energy. Hopefully a solution will arrive soon.
(nb if this is a duplicate, not intentional but I kept losing comments before posting)

Mar 22, 2015 6:29 AM ET

Response to Janice Bell
by Martin Holladay

I understand your frustration. A couple of points:

1. If you really want to install a solar thermal system -- and considering the very real issues with maintenance in your climate, I don't recommend it -- I don't see why you would need a 1,500-gallon tank. Most residential systems have tanks with a capacity of 120 gallons or less. The only reason to install a 1,500-gallon tank would be for a space heating system.

2. The fact that some utilities are not accepting new net-metering customers is, indeed, frustrating. My advice is to plan a roof that is capable of receiving a PV array, and sit tight. Within a few years (I predict), one of two things will happen: (a) Due to political pressure, utilities will allow homeowners to connect their PV systems to the grid, and will provide credits that make sense, or (b) The price of batteries will drop to the point where grid connection is irrelevant.

Mar 22, 2015 10:47 PM ET

Edited Mar 22, 2015 10:51 PM ET.

Response to Martin Holladay
by Graham Irwin

To clarify, I was not seeking your blessing nor you approval to pursue a vegetarian or a car-free lifestyle, I was questioning whether these approaches are any less valid "credits" to claim against water heating than PV electricity supplied to the grid at an unrelated time.

To heat water with PV, I make it that you'll need 2x the area in PV of the solar thermal panels and a heat pump with high enough capacity to utilize the output of the PV panels at an average COP of 2. This system would be costed against a solar thermal system with electric resistance backup because times with low solar energy are likely to coincide with minimal heat pump performance (though this should be verified.)

If space were constrained, one might conclude that the less efficient but more versatile output of the PV panels is better sold to the grid or used to charge electric cars or the like and use it in conjunction with solar thermal for its space efficiency.

In any case, you face a steep uphill climb against the laws of thermodynamics to convert heat into electricity and back into heat again, all to eliminate an incremental bit of plumbing, but that's just physics, not "magic."

Mar 23, 2015 5:50 AM ET

Response to Graham Irwin
by Martin Holladay

Your description of the needed comparison -- comparing a PV array and a heat-pump water heater to a solar thermal system with electric-resistance backup -- is of course exactly the comparison that was made in my article. You're right that the solar thermal system requires a smaller area; but as my comparison shows, the PV equipment costs less and is more dependable. So homes without much area that can be devoted to solar panels may prefer to install a solar thermal system, as you point out, although such a system will cost more and be less dependable.

Apr 6, 2015 7:02 PM ET

Is solar thermal for pools dead too?
by Tom McReynolds

Interesting article.Thanks for giving me a new perspective!

Does this argument scale up for pool heating as well? Please forgive me if this has been answered already, I didn't see an answer in the comments.

I also wonder if the type of thermal makes a difference. I'm using evacuated tubes and I have enough heat for hot water year round and the pool > 80 for about 6 months of the year. I also have a lot of PV. I ask because evacuated tube and flat panel costs are converging, and evac tubes gives you more independence from local temperature.

But it never occurred to me to price it against pure PV plus heat pumps though. I know there are heat pump pool heaters, so maybe efficiency scales up and gives the same answer.

I live near San Francisco, so I have an unfair advantage on climate. :-)

Apr 7, 2015 4:06 AM ET

Response to Tom McReynolds
by Martin Holladay

The economics of heating swimming pools with solar thermal collectors is entirely different from the economics of heating domestic hot water with solar thermal collectors, due to the fact that the target temperature for swimming pools (as low as 80 degrees, as you point out) is generally much lower than the target temperature for domestic hot water (120 to 140 degrees).

The type of solar collector used to heat swimming pools is usually a simple low-temperature collector consisting of unglazed black plastic tubing. Because this type of collector is cheap, and because the desired water temperature is relatively low, this type of collector is cost-effective -- compared to heating pool water with fossil fuels, of course, but not compared to diving into the unheated swimming pool regardless of the water's temperature.

May 1, 2015 12:42 PM ET

Problem with your math
by no spam

Your PV system is making a wrong assumption and is undersized. If you used your numbers in an off grid situation you would be taking a lot of cold showers.

I have a HPWH and have measured that it uses 3-5kWhr of electricity IN ONE HOUR to heat a tank of water during normal daily usage. I am also averaging 10-15kWhr/day electricity to heat water for a family of 4. You seem to assume that you can use all day to heat the water. A 0.57kW system wouldn't even work.

Look at the math - assuming somebody in Toronto, they can get an "average" 3kWhr of electricity generated per day per kW of solar panels used. So, at least, I would need a 2kW system = 2000Wx$4/W=$8000 system. In reality you will find that you need a 5kW system with batteries to early keep up with your water heaters demand - the would cost $20,000. PV is not cost effective for heating water as its efficiencies are way to low.

May 1, 2015 1:04 PM ET

Response to No Spam
by Martin Holladay

No Spam,
You are correct that it makes no sense to use an off-grid PV system to heat water. I never advised off-grid homeowners to do that.

The economics of grid-connected PV systems are entirely different from off-grid systems, however. My math assumes a grid-connected system, and my math is accurate.

I have lived in an off-grid house for 40 years. I heat my water with a solar thermal system in the summer and a thermosyphon loop off of my wood stove during the winter. Both systems use a propane-fired tankless water heater for backup.

Jul 20, 2015 7:11 AM ET

Run a similar comparison for GSHP?
by Clay Whitenack

Hey Martin. I really enjoy these mental exercises. Could you do a similar comparison between GSHP and PV powered ASHP? I have read your articles in the past where you say that GSHP use a lot of electricity compared to ASHP and in one article you flat out recommend not using GSHP. If I take the money I would save in not installing a GSHP plus the extra cost of operating one each year, how much money would I have to buy PV, and could that come close to generating the energy needed to power an ASHP?

Jul 20, 2015 7:26 AM ET

Response to Clay Whitenack
by Martin Holladay

Many energy consultants have performed the analysis you suggest, and air-source heat pumps (usually ductless minisplits) plus PV always beat a ground-source heat pump -- whether you are talking about a small residential installation or a larger commercial/institutional installation.

One such consultant is Andy Shapiro, who designed an extremely energy-efficient gymnasium (the Field House) for the Putney School in Vermont. Here is a link to an article that describes Andy's analysis:

"The decision to utilize heat pump electric heating was made based on the decision to achieve net-zero performance with on-site electricity generation. A spreadsheet analysis was created to compare the annual performance of a VRV air-source heat pump system to a ground-source heat pump system. This model predicted an annual coefficient of performance (ACOP) of 2.3 for the air-source heat pump (inclusive of indoor fan power), based on Concord, NH weather. This compares to an assumed (based on experience) ACOP of 3.5 for the ground-source heat-pump (including ground water and distribution energy). The size of the system was calculated to be approximately 28 heating tons.

"For the ground-source system the cost of the boreholes was estimated to be $106,000 and the cost of the water-source heat pump system inside the building was estimated to be $200,000, for a total system cost of $306,000. The cost of the air-source system was estimated to be $200,000. The cost of the added PV array required to offset the added electrical load of the air-source system compared to the ground-source system was $38,000, based on the additional electric consumption of 6,400 kWh/yr. The calculated net savings by going with air-source heat pumps instead of ground-source heat pumps was $68,000 including the added PV area required. Thus, the decision was made to proceed with design of the air-source system."

Sep 10, 2015 12:59 PM ET

PV versus SHW System pricing
by Adam Coffman

I have been planning for two years now to install a solar hot water system, and just now stumbled across this website. There is some excellent information here, but it seems to me that the cost of a solar hot water system has been grossly overstated. I have found it difficult to price out systems, as most are intended to be installed by companies dealing in these systems and I am planning on installing one myself. I have found the following systems:
The one that most fits my needs would be the 30 tube/80 gallon. The cost of the system is $2,499. Add in the turnkey option (for all the needed hoses, fittings, etc.) for another $395. The site claims to have a good deal on shipping, but I haven't checked into it yet. I will assume for now that it is around $200, which brings the total cost to $3100. The collectors are eligible for the 30% tax credit, so that brings the cost down to $2170. As hard as it has been to find any other system prices for SHW, it is just as difficult for PV. Can anyone tell me if I can purcahse a PV system that I can install myself for anywhere close to the SHW price? I am not committed to SHW, so if PV can be done as cheap I would definitely like more information on where to purchase a PV system. One drawback to PV that I have found in my area is that the utiltiy company makes people jump through hoops to get them tied into the grid.

Sep 10, 2015 1:48 PM ET

Response to Adam Coffman
by Martin Holladay

Just because you can purchase most of the hardware needed to install a solar thermal system for $2,499 + $395 = $ 3,894, doesn't mean that the system can be installed for that price. If a solar contractor has to come to your house, discuss the installation, get up on your roof to install the collectors, use a licensed plumber to install the equipment, and provide a warranty (and maybe have a callback occasionally), that work will cost $7,000 to $9,000.

Whether or not you can install a PV system on your house depends on (a) whether or not you are a licensed electrician, and (b) your local building regulations. Call up your local building department to find out what regulations apply in your town.

Sep 10, 2015 3:37 PM ET

Response to Martin Holladay
by Adam Coffman

I work with heat transfer fluids on a daily basis. The plumbing is not an issue for me. I may need to obtain a permit to install the system, but I have been unable to determine that so far. I will need to check with the township. I am certain that I would not need to hire an installer. What I am looking for is information on PV systems. I can easily do the electrical work myself. It just needs to be inspected by an electrician. I am planning on using a lower roof on my house (it is the most robust as the main roof was built in the late 1800's), but that roof is only 12' x 8', so it may not be large enough anyway for PV.

Sep 10, 2015 3:48 PM ET

Response to Adam Coffman
by Martin Holladay

If you can easily do the electrical work yourself, then you are all set.

If you don't know enough to do the work yourself, consult an electrician.

Sep 10, 2015 3:59 PM ET

Edited Sep 10, 2015 4:01 PM ET.

by Malcolm Taylor

When trying to compare alternatives it simply muddies the waters to include DIY as part of the discussions. In individual cases of course it may influence the costs significantly, but that should in no way alter the apples to apples comparison of two technologies.

Sep 10, 2015 8:45 PM ET

Edited Sep 11, 2015 9:46 AM ET.

DIY comparison
by Charlie Sullivan

DIY grid tie solar kit prices are about $2/W, and the cost is probably similar if you buy the pieces. Permitting is more complex because you need the utility involved, but let's put that aside for now. Using $2/W and Martin's estimates of 2.05 kW or 0.68 kW for resistance or heat pump water heating, you need $4.1k or $1.36k worth of PV hardware. The latter is cheaper even after you buy a heat pump water heater, especially if you opt for the GE unit at the low end of the cost range at $1k. And it's easier to fit the smaller array on your small roof, if you go with the HPWH. That appears to come in a little cheaper than your $3100 solar thermal option, but it's not the slam dunk win for PV that Martin finds for professional installation. So it seems to me that you could go with whichever seems more manageable in terms of your DIY skills and the permitting hassles.

Here's an example of a 2.04 kW PV kit for $3900, or $4260 with a mounting kit

For lower power and smaller roof area, you could buy the components separately, or ask them for a custom kit with 3 panels instead of 8, and 3 microinverters.

Edit: reply to Malcolm's comment above. Yes, it is confusing to include both DIY and professional costs in the same discussion. But which one is the "right" one to consider depends on the individual homeowner's plan. What's important is to be clear about which comparison you are making and to avoid overgeneralizing the results. So to anyone considering a professional installation, ignore my numbers above and go to the top of the page to Martin's comparison.

Sep 28, 2015 9:20 PM ET

Solar Thermal BS
by jp dortmunder

I am sick of this nonsense. I have spent 30 years in the solar business and every company's sales force looks like a scene from the Saprano's. It's a giant ripoff scam and all of you morons are using the ridiculously inflated prices that keeps these bastards wearing Rolex watches and driving expensive cars. The approximate equipment cost of your 2-4x8 panel, 120 gallon tank system is $3000. A good two man crew will install this system in about five hours and be paid about 300 bucks. So your installed price is $9000. You stupid dishonest bastards.

Sep 28, 2015 9:36 PM ET

Martin Holiday
by jp dortmunder

$2499+$395=$ 2894 not $3894. The installation will cost the contractor about five hundred bucks. This industry is ruined by moronic thieves. Six grand is a one hundred percent mark up. Let alone $9000. It's a friggin gold mine for these vermin.
Adam, don't go with vacuum tubes. They are unreliable and have a high rate of failure. They lose their vacuum.

Sep 30, 2015 4:05 AM ET

Completely different calculation for India
by Gurudatta Sardessai

I am amazed at the calculations. Here in India we use Solar Thermal extensively. If we use a 2 Panel Solar thermal that gives us between 200 to 250 Litres of hot water per day. (around 60 gallons). The same setup gives 63% of 44 gallons (about 105 litres in the climate that you have assumed in your calculation).
All these calculations apart.; a two collector system with insulated tank with a 10-15 year life costs INR 50,000 versus a 2 KWp PV (grid fed - although that is not yet available for us everywhere) which cost upwards of INR 200,000 (4 times that of a thermal system).
As for space - a 2 panel thermal system will need about 1/4 the space for the 8 numbers of 250Wp panels that a 2 KWp system will need.

Sep 30, 2015 8:19 AM ET

Latitude & climate matter! (local economies do too)
by D Dorsett

india is blessed with a lower latitude, which results much higher insolation than in the US. Only in the desert southwest (the sunniest part of the US, outside of Hawaii), does the sunshine exceed the average in India. Most of the populated US receives much less. In my city in the northeastern US there is a large number of immigrants from Ghana, some of whom complain about how dim the sun seems to be here.


Also, in most of India there is no risk of the solar collector or pipes freezing, whereas in most of the US systems have to be designed to manage those issues, which the solar thermal system more complex & expensive. Photovoltaic (PV) panels don't have those problems, though it's important to mount solar panels (any type) so that the snow can slide off easily.

The cost of PV is falling everywhere. Under the current Indian administration policies the size of the industry in India will grow by more than an order of magnitude in the next five years, which will drive the installed price of rooftop systems down rapidly. Whatever else people might think of Narendra Modi, he is not a puppet of Coal India, and he managed to bring a LOT of solar power to Gujarat while he governed there. (But it was primarily large utility scale arrays, not residential rooftops.) Rooftop solar pricing should drop by half in the next five years, but it's not clear whether heat pump water heaters will become cheap & available in India any time soon.

May 9, 2016 3:45 PM ET

HPWH Assumption COP 3 is Too Optimistic by Double
by Tania G

Martin's math has a couple fatal flaws.

1. Martin assumes Rosenblum's is correct that a typical HPWH has a COP 3.0 annual average, this is a significant error that cannot be justified. The NREL testing found the average HPWH unit was under COP 2 with averages closer to 1.6. Add in the Canadian CHMC study on the Parasitic load a Heat Pump Water Heater puts on the Furnace, and the net gain was close to 1.2. The size of a PV system is more realistically in the range of 1.5 kw or higher.

2. The installed cost of 1 kw PV is about $6000-7500 here. It is a mistake to use any $x/watt estimate because 4 panels, roof hardware, inverter, and install starts at $6k to show up. Try the math at $6-7/kw and for 1-1.5 kw and the article would be more credible.

3. Energy Factor used for HPWH is not the same as COP used for Air Source Heat Pump. Even ASHP and GSHP differ on COP, leaving out the fans or pumps to boost the numbers. The Sun Pump solar hot water heater has closer to field tested COP than in Lab calculated that omits required electrical consumption.

May 9, 2016 4:36 PM ET

Response to Tania G
by Martin Holladay

As I wrote in my article, "Using the information in this article, GBA readers can perform their own calculations." I clearly stated my assumptions and the basis for these assumptions.

Everyone's costs are different. COP data and equipment specs are both moving targets. Feel free to plug in your own numbers and reach your own conclusions.

May 9, 2016 5:33 PM ET

Residential PV in the US is ~ $3.50/watt, not $6-7/watt
by Dana Dorsett

This is consistent with the installed cost bids I've reviewed for 4-15 kw systems in MA in the past 12 months. There are locations in the US where the installed cost is well below that average, and locations where it's well above, but $7500 /1kw would be more than 2x the US average. Even for a sub-2kw system that's quite an up-charge!

Out of curiosity, where is "" in the statement "The installed cost of 1 kw PV is about $6000-7500 here." ?

May 9, 2016 6:01 PM ET

Second response to Tania G
by Martin Holladay

As I wrote in the article, "PV plus electric-resistance approach is about 25% cheaper than the solar thermal route." What this means is that my argument is valid even if you totally discount the idea of using a heat-pump water heater.

But, as I said in my earlier response, perform your own calculations and choose the equipment you prefer.

May 9, 2016 6:40 PM ET

I am afraid I passed up on
by Jonathan Lawrence CZ 4A New Jersey

I am afraid I passed up on the deal of the century. I got a quote back in December of 2012 from a major solar company for a 10KW DC system for $17,600. So $1.76/W DC. That was a prepay plan, so I believe solar company got the tax incentives and SREC income. That came out to $.076/kwh over the 20-year life of the contract. I will be getting a quote from them in the near future for my upcoming PH.

I also got a quote from a regional company at the same time for a 7.2 KW DC system for $24.4k after ITC, or $3.4/W DC. They estimated $19k of SREC income over 15 years and a 20 year production of 160,712 kwh, for a cost/kwh of $.031. Even if the SREC's went to $0, the cost would be about $.15/kwh, which compares to my cost at the time of $.167/kwh.

A big part of the marketing pitch by the salesmen at the time was the ability to lock in the rate now because the utility rates were going to increase 4% per year. I have TOU now, but just checking the tariff rate in effect for my area, averaging the summer and winter rates, the $/kwh is now - $.167, the same as 2.5 year ago.

Dollar and sense aside, if I had pulled the trigger, I would have reduced my carbon footprint quite a bit. Won't make that mistake on the next house.

May 9, 2016 8:14 PM ET

RECs and carbon footprint
by Charlie Sullivan

Jonathan, if you had sold the RECs, you would have sold the right to claim you reduced your carbon footprint. You would have given someone else buying renewable certified electricity from the grid the opportunity to reduce their footprint, which is a good thing, and you would have saved money, which is also a good thing, especially given that (based on your other comments) you tend to invest that money in worthwhile ways. So I'm not against it, but I think that people who participate in the REC market should respect it.

May 9, 2016 8:41 PM ET

Edited May 9, 2016 8:41 PM ET.

Charlie,The SREC pricing is
by Jonathan Lawrence CZ 4A New Jersey


The SREC pricing is NJ is currently $280/MWH. That will help to pay for those triple-pane windows, batteries and other PH improvements. So this could be a win/win where I reduce my energy consumption by 80% and the utility gets to add solar generation to its network to be bought by someone else. However, it will be interesting to see what the utility does when my energy flow is one way 99% of the time. I am sure they will send me a nasty gram eventually. I might end up selling my power directly to my neighbor.

Jul 1, 2016 3:28 AM ET

Edited Jul 1, 2016 3:50 AM ET.

Historic Thread
by Kevin Dickson, MSME

I was pleased to see comments above from well-known experts in solar thermal: Tom Gocze, Tom Lane, and Fortunat Mueller.

Another person worth mentioning is Jay Burch of NREL, who spent many years searching for low cost high reliability solar DHW. He never really found it, and is now retired.

In 2012, I built a prototype solar sytem for my house using some of Fortunat's basic research, low cost Chinese evacuated tube collectors, and my 33 years of experience. The goal was a sub $3000 domestic water heating system. It worked pretty well for a couple years but alas it did freeze and break. I still think that cost goal could be met with an intense failure-mode-elimination-effort , but I'm also convinced it's not worth it, PV has won.

Jul 21, 2016 3:40 PM ET

PV & Solar H2O
by Kenneth Sayers

I have a 2.1 kW PV system and an 82 gal. Rheem solar hot water heater with one collector, using a closed loop system. The whole thing, together, cost me less than $8,000 after rebates. The collector provides me with 82 gallons of water at about 160°F about 85% of the time. This allows the PV to assist with all my other appliances, TV, computer, and heat pump. I keep the house around 76°F and my electric bill runs me, on the average, less than $70/month. The electric company charges me 19¢/kWh and pays me 5¢/kWh for what I generate in excess of what I use.

I would have to say that it works for me. I must say, however, that my 1600 sq foot home is really energy efficient.

Jun 19, 2017 3:30 PM ET

Counter Points
by j moore

The article incorrectly inflates the cost of the solar thermal system. Adding the cost of a separate electric resistance HW heater to the system cost is incorrect. An electric element backup is normally included in the solar hot water tank at no extra cost. Most systems have no electric backup but simply "pre heat" an existing gas, oil, or propane fired HW tank thus reducing use
of inefficient fossil fuel-based HW systems.

The analysis disregards current incentives. For the PV options, the tank cost would not qualify for a 30% federal tax credit, while with solar HW, the tank does qualify, since it is integral to the system and stores solar heated water. Also, rebates vary by state and need to be accounted for. In several states, rebates are very generous for solar thermal systems- up to 40% of the system cost!

Roof space required for a PV option of equal production is 2-3 times that required for solar hot water collectors. That extra space is not always available. For maximum energy harvesting, best use of roof space is a combination PV panels and solar thermal collectors.

No way a PV installer is going to install a tiny 5-6 panel PV system for
$3.75/Watt. Expanding an existing PV design to accommodate HW load can make sense, but again roof space can be an issue.

The jury is still out on heat pump hot water heaters in northern climes. They rarely are in heat pump mode in cold basements in the winter and if in heated spaces, they make space heating systems work harder and burn more fuel. Also they are loud, have unproven life, and require an expensive refrigeration expert for repairs and maintenance. Solar thermal systems consist of robust hardware and if installed and maintained properly, have consistently demonstrated system lives of 20-30 years.

Jun 19, 2017 4:00 PM ET

Edited Jun 19, 2017 4:44 PM ET.

Response to J. Moore
by Martin Holladay

I compared the cost of the equipment required for a solar thermal domestic hot water system with the cost of the equipment required for a heat-pump water heater.

You are apparently talking about a one-tank solar thermal system -- a type of system with many disadvantages, which is why such a system is rarely installed. A quality system is a two-tank system. (You need an electric-resistance backup heater because the sun doesn't always shine.) The main problem with a one-tank system is inefficiency -- the electric resistance element keeps the water in the tank warm, so there is never any cold water available to send to the solar collectors. (The efficiency of the solar collectors depends on the delta-T -- the colder the water you send to the collectors, the higher the efficiency of the system.)

You're correct that I disregarded incentives in my analysis. (That point was clearly stated in the article: "The price comparisons made in this article do not include any incentives, rebates, or tax credits.") Incentives vary from state to state and from country to country, and it's possible to cherry-pick a location to prove almost any point you want to prove. In many locations, contrary to your statement, incentives and rebates have long been in place for purchasers of heat-pump water heaters.

You are absolutely correct that a solar thermal system usually takes less area than a PV system of comparable output. If your roof is tiny, you should go ahead and install a solar thermal system (if you can afford one).

The listed price for a PV installation of $3.75/watt is, indeed, obsolete -- but the price change is in the opposite direction from your implication. PV is cheaper, not more expensive, than noted in this article. Here at GBA, we're getting many reports from U.S. residents who have installed systems for $2.75/watt, which only makes my argument stronger.

There is no need to install a tiny PV system. PV is so cheap, and the return on investment is so attractive, that you should install the biggest PV system you can afford. This investment will outperform any stock portfolio in your retirement account.

If your point is that there is a minimum reasonable size for a PV system, you're right. Depending on your needs and expectations, that minimum size might be 1 kW, 2 kW, or 3 kW. Readers who conclude that an $8,000 PV system is unaffordable won't be able to buy one, of course -- but they won't be able to afford an $8,000 solar thermal system, either.

In general, I've heard of more maintenance issues with solar thermal systems than with heat-pump water heaters. It's possible to have good luck or bad luck with either type of system. For readers who are wary of heat-pump water heaters, I recommend that you re-read my article, and note my calculations on an electric-resistance water heater coupled with a PV system.

Jun 19, 2017 5:17 PM ET

What a difference 30 months can make (in PV pricing)
by Dana Dorsett

The average cost of sub 10KW rooftop PV in the US is now closer to $3/watt (all in, no subsidy), than the referenced $3.74/watt. ( ) In some local markets it's closing in on $2/watt.

The value/flexibility of PV output is greater than the value of thermal output of solar thermal of equivalent rooftop real estate too.

The EF efficiency of heat pump water heaters has now climbed into the low 3s, up from the low 2s back when that analysis was done, even as the average retail pricing has slowly declined. (eg: ) The improved water heater efficiency reduces the size of the PV array necessary to support it by about a third, so instead of 570-680 watts in the examples used in the article, it's now only takes ~380-450 watts of panel.

In Massachusetts a 55 gallon or smaller HPWH with an EF of 2.3 or greater qualifies for a $750 rebate, down from $1000 a couple of years ago, to reflect the drop in the hardware pricing. For bigger tanks the rebate is smaller (probably reflecting the projected higher income of those buying larger water heaters.)

Mar 19, 2018 1:14 PM ET

... unless you have an attached solar Greenhouse.
by Evan Ravitz

If you just run say 50 feet of black hose through your attached solar Greenhouse to the input of your water heater or pump, you'll save plenty. In many places, no freeze protection or transparent collector cover is needed, so the real cost is the black hose.

As to maintenance, with a simple thermosiphon design, that should be limited to water leakage. But in a greenhouse with a dirt floor, that should be a benefit: free plant watering!

Apr 20, 2018 9:36 PM ET

I know you're right, but...
by Domenico Perrella


I know you're right that the most cost-effective way to use the sun for water heating or space heating would be to buy some PV panels and use them to run a heat pump or two, but the sun is so nice and toasty warm that there is an intuitive belief that there must be some way to use it that isn't just flushing money down the toilet relative to some other means that would reduce fossil energy consumption more at less cost.

So, here's my whatabout question. Whatabout if a person were about to replace his composite shingle roof with a metal one, was expecting to use metal purlins on top of continuous insulation outside the roof anyway, knew of a place that advertises (and maybe actually sells... maybe) purlins with a little rounded groove to hold a pex line up against the metal roof to be warmed by it. Is there maybe some tiny chance that it would be economically rational to set up a couple of zones on different areas of the roof where the heat during the day could be used to warm glycol that could be used for water preheating, maybe a little subfloor heating in winter, and (bonus) subfloor cooling on summer nights to chill the house down to minimize AC use the next day.

Yes, I know that the water wouldn't be as hot as with a well-designed thermal collector, that the system would require at least a couple of pumps, several automated valves, a big insulated tank, and the homeowner (OK, it's me we're talking about) would have to program a controller himself (let's say I consider that challenge a bonus, even though I don't really have time).

Now comes the part when you explain that the roof would be cold when space heating was needed, that even in summer, it probably wouldn't get water quite warm enough for domestic hot water use, that parasitic losses from the pumps would exceed any energy savings, and that only crazy people try to chill there houses to 65 at night in the summer to avoid running the AC the next day. However, we already try this just using open windows and fans and it really does seem to minimize cooling except on a handful of days when it never gets that cool at night. And, no, we don't ever really manage to get the house down to 65 at night in the summer, but with a stonking big radiator on the roof, I'll bet we could.

Please ignore the fact that my house isn't plumbed for subfloor heating and cooling and that it's a bad choice for me to pay to install that when mini splits work better. Hypothetically, if I already had the subfloor heating set-up, could it maybe make a little sense?

I'm not actually going to do this if it doesn't make sense, but I really wish it did because the idea just appeals to me at a non-rational level. I think the space heating probably makes no sense because the roof can't get that warm on the days when I'd most want heat. Also, once I get the house decently insulated, I won't need heat on the days when the roof does get fairly warm. Water heating might work during the warmer half of the year, but not necessarily at a reasonable cost. Also, I feel like the nighttime cooling during the summer could work and I like it cold when I sleep, but would it justify the cost? I guess not.

Facts: Northern California (East Bay), much more heating than cooling, roof gets pretty much sun, but only a couple of areas almost never get shaded by trees, and our sunny hours are somewhat limited by being in the bottom of a valley.

Apr 21, 2018 7:10 AM ET

Response to Domenico Perrella
by Martin Holladay

This is an old story. Several companies have marketed systems like the one you describe. In all cases, the value of the heat gathered by these systems is too low to justify the high cost of the equipment required to collect the heat.

Surprisingly, this experiment is engaged in every 5 years by a new enthusiast. At the end of the experiment, the enthusiast adds up the cost and headaches and realizes that very little heat was gathered, and the cost is high.

I suppose you can run the experiment one more time, expecting different results, if you want. When you are all done, report back.

May 8, 2018 10:36 PM ET

Thanks, Martin
by Domenico Perrella


Apologies for the belated response and thank you for your response to my question.

You know, I really wrote that post with the idea that you would save the day by talking me off that ledge. It is really hard for me to overcome the intuitive belief that all the heat I can feel from the sun must be usable in some way to heat my house. The tricky part is that most of the nice warm sunlight occurs in the months when I don't need heat and there's a limit to how hot an exposed metal roof can get if it's 45 degrees out. That's an advantage that an enclosed collector has over a big exposed roof, although it has higher costs per square foot to go along with that advantage.

But, what about my effort to avoid the results of prior experiments? Can we also totally dismiss the ability to chill the house during summer nights? Even on a warm day a lot of heat radiates out into space from a warm object like a roof that is being kept warm by cooling something large and warm, like a slab (if my house were built on one) or maybe cooling the air in the house. I have to believe that, with a little insulation in a house without many windows exposed to summer solar heating, chilling the house to about 65 at night would keep it below 77 or 78 all day, which I can live with. But I guess I can't get much of a cost offset with that since I've already admitted I don't spend a lot on cooling.

Well, you've rained on my parade, but that's what I actually wanted you to do.

Thanks for being a spoilsport.


May 9, 2018 6:32 AM ET

Response to Domenico Perrella
by Martin Holladay

If outdoor temperatures are cool at night, and you want to take advantage of those temperatures to cool your house, use a whole-house fan. For more information, see Fans in the Attic.

Register for a free account and join the conversation

Get a free account and join the conversation!
Become a GBA PRO!