heating hot water with pv panels

[Ampster]

..........

Do you have anything more to offer or add to the conversation beyond pointing to what others have to say/write ? What do you know of the subject ?

I have 3 HPWHs installed in buildings I own and they have paid back their cost in less than 4 years compared to a resistive electric water heater. What is your real world experience using HPWH?

No doubt a HPWH has to work harder in cold climates. It doesn't take rocket science to conclude that the cost effectiveness will be lower in a colder climate than in a warmer climate. A HPWH is still more efficient than other forms of electric water heating. The simple concept that most people can understand is that it is less expensive to move heat than to create it when using electricity as the heat source. Natural Gas, when available, is still the best return..

[nwdiver]

I have a NYLE (built in Maine so they know cold) external HPWH and it is sad they left the home market. They used a different refrigerant that was a little more suitable for the colder climate. I have a couple of engineering studies on HPWH and just don't believe COP of much over 2. Refrigeration technology hasn't changed that much. It is all about the fine print and how far you want to stretch the truth.

Here's a test of the Sanden HPWH. They measured a range of COP from 2.1@17F to 5.0@95F.

Comparing a HPWH to nat gas is tricky especially if you have solar and can't roll your surplus at full retail month to month which is what I have. Even if the gas was free it was still cheaper for me to use a HPWH just because of the $15/mo connection fee to have gas service. My HPWH uses <400kWh/yr and I get <$0.03/kWh for my excess generation. Plus it helps cool and dehumidify my garage.

[J.P.M.]

I have 3 HPWHs installed in buildings I own and they have paid back their cost in less than 4 years compared to a resistive electric water heater. What is your real world experience using HPWH?

No doubt a HPWH has to work harder in cold climates. It doesn't take rocket science to conclude that the cost effectiveness will be lower in a colder climate than in a warmer climate. A HPWH is still more efficient than other forms of electric water heating. The simple concept that most people can understand is that it is less expensive to move heat than to create it when using electricity as the heat source. Natural Gas, when available, is still the best return..

Thank you for the response. I'll take that as a no answer to my question.

As to your question: My experience with HPWR's is probably more than your experience in dealing with heat pump systems in cold climates. I did however fit a prir residence in Buffalo with a HP HVAC system. Turns out it was OK for the shoulder seasons but used more energy at ~ 2X the cost of CH4 fuel in the cold months. Bad move. Did it before I became an engineer.

[nwdiver]

Thank you for the response. I'll take that as a no answer to my question.

As to your question: My experience with HPWR's is probably more than your experience in dealing with heat pump systems in cold climates. I did however fit a prir residence in Buffalo with a HP HVAC system. Turns out it was OK for the shoulder seasons but used more energy at ~ 2X the cost of CH4 fuel in the cold months. Bad move. Did it before I became an engineer.

How long ago? New HPWHs are significantly better than older ones.

[PNPmacnab]

Getting back to actual PV solar water heating, it works for me. This is a chart of about 11 days of water heating taking showers, dishwasher, etc. Temps drop down overnight to about 107 and this thermostat (likely because of litigation threats) can only go up to 127F even though the data sheet says 150. It normally shuts off about 10am. It doesn't prove anything except it meets my needs. The one thing that really shocked me is that it starts heating about 7am. And this energy is from the same panels that recharge my battery. Data was taken by an Elitech RC-5 which for a cheap data logger packs a lot of features and every data point on the chart can be accessed just moving the mouse. I think I sampled every 5 minutes. morning_water.jpg

[J.P.M.]

How long ago? New HPWHs are significantly better than older ones.

Well, that was in the '70's and it wasn't with HPWH which were not available in any real sense for residential use at the time, and haven't been around that long even now, but I retrofit a heat pump as an adjunct to a CH4 fueled residential furnace on the next door property I owned.

I was responding to Ampster's comment and what read to me like his acknowledgement about HPWR being less efficient (and so probably implied less cost effective) in cold(er) climates, and also to his question about my experience with HPWRs as a matter of information and clarification. I also suspect his experience with HVAC equipment in cold climates is not as extensive as might be needed to discuss the issue in an informed way.

As for improvements in air-air HP systems, I'm pretty up to date on what's happened and what may be on the horizon. Seems to me most of the improvements of the last 20-30 yrs. or so come from:
- Improvements in the Thermodynamic properties of the working fluids.
- More flexibility in the operating speeds and working fluid handling (for example, multispeed compressors and air handlers).
- Some improvements in mechanical operation (scroll compressors, etc.)
- And what may be the biggest improvement: Massaging the hype and sales pitch to make it easier for users to think the actual improvements produce more efficient operation with the implied or vaguely promised associated improvements in unit cost effectiveness. All the bogus crap about EER/SEER comes to mind. Seems to me that's a better tool for B.S'ing users than it is for users who don't understand any of it, but rely on the hype. Seems to me a lot more objective reality would be conveyed by simply sticking to COP. SEER's of 20 - 30 are mostly B.S. anyway as is the whole slippery EER/SEER scheme.

As an example of uselessness: York's use of a smiley/frowny icon on the their monitor showing OK/trouble operation as an improvement.

One thing that doesn't seem to have improved and that mfgs. don't seem to talk about much, and as best as I can tell isn't accounted for in anywhere is the method and energy used to get the frost off a HP heat exchanger when the coil temp. drops below the frost/dew point temp. in the winter.

FWIW and IMO only, most of those improvements tend to benefit operation in cooling mode. The lower COP's as f(outdoor amb. temp.) in heating mode make it more difficult to obtain the same cost effectiveness as is possible with nat. gas for heating. The same applies for HPWR.

HPWH are similar in how the heat is supplied, but have their own and separate considerations, a big one being heat exchanger fouling.

And too, in colder climates and where the heat source for the HPWR is the indoor air, that supplied heat must eventually come from some other source, that source usually and mostly being the HVAC equipment.

Also, the kiss principle tends to work against HPWH when compared to the dumb, plain old, but mostly tried/true/debugged tank type water heater. Not sexy and few bells/whistles/icons that smile, but if gas fired, safe, cheap to buy and cheap to operate over the long haul.

When all is said/done, for my thinking and money, HPWR for residential use in cold(er) climates are probably not as practical, trouble free or as cost effective as peddlers would have us all believe, particularly if nat. gas is available as a fuel source.

Take what you want of the above. Scrap the rest.

[nwdiver]

Also, the kiss principle tends to work against HPWH when compared to the dumb, plain old, but mostly tried/true/debugged tank type water heater. Not sexy and few bells/whistles/icons that smile, but if gas fired, safe, cheap to buy and cheap to operate over the long haul.

When all is said/done, for my thinking and money, HPWR for residential use in cold(er) climates are probably not as practical, trouble free or as cost effective as peddlers would have us all believe, particularly if nat. gas is available as a fuel source.

Except in this case the end result is vastly different. For most people in most places for most of the year a HPWH is going to cut their energy use for heating water by ~70%. It really doesn't make much sense to waste energy heating an element when there's so much ambient energy when it's >40F.

There's really nothing exotic or complicated about a HPWH... it's just a refrigerator working backward.

In terms of cost effectiveness a HPWH is going to be the ~3rd most cost effective upgrade ~roughly on par with DIY solar and ~twice as cost effective as turn-key.

[J.P.M.]

except in this case the end result is vastly different. For most people in most places for most of the year a hpwh is going to cut their energy use for heating water by ~70%. It really doesn't make much sense to waste energy heating an element when there's so much ambient energy when it's >40f.

70% of what ? Over nat. Gas. Which is my point, i kind of doubt it. The conversation was relating to cold climates.

an example of why i have my doubts about 70 your 70$ savings: I'm not in most places. I currently live in a mostly non freezing climate in inland san diego county. The coldest 31 day ave. Min. Low temp. In my 11 yrs. Here has been 42.4 f. It's about as temperate as it gets and about as ideal a climate for heat pumps as it gets.
Currently, (06/01/2019 t.o.u. Rate sheet 31819 - e) sdg & e super off peak t.o.u. Electricity rate is ~ $0.35/kwh, summer and winter. That means a therm of electrically provided heat used with a hpwh cop of 3 will cost ~ $0.35*((100,000/3,412)/3 = $3.42 per therm delivered to the load. The proposed incremental cost of a therm (100,000 btu) of nat. Gas from sdg & e for residential customers currently has an incremental cost of ~ $1.71/therm (it's currently ~ $1.51/therm). Burned at, say, 70 % efficiency that comes to ~ $1.71/.7 = $ 2.44/therm delivered to the application. At the $ rates used, the year round cop of the heat pump would need to be ~ 4.1 to begin to break even with nat. Gas, even before the increased costs of equipment and likely increased maint. Costs of the hpwh are taken into account. Also, to increase the cop to 4.1would probably require daytime operation because nite temps (remember the super off peak rate thing) are lower -->>> lower cop's.
So, $3.42 vs. $2.44 for heating cost for a heat pump vs. Nat. Gas in what's thought of as a pretty benign climate using published super off peak rates for the electricity and (proposed higher) published rates for nat. Gas, and delivered costs accounting for cop's and thermal combustion(in)efficiencies.

there's really nothing exotic or complicated about a hpwh... It's just a refrigerator working backward.

Yea, so is turning a window shaker around and using it as a heat pump. So ?

in terms of cost effectiveness a hpwh is going to be the ~3rd most cost effective upgrade ~roughly on par with diy solar and ~twice as cost effective as turn-key.

but in terms of the application - supplying dhw service to a residence - both in terms of initial cost of installation, and fuel cost, not to mention the kiss advantages of using simple tech., it's hard to beat nat. Gas if available, even in what's probably an ideal climate for hpwh use.

j.p.m.

[nwdiver]

j.p.m.

$0.35/kWh is at best a disingenuous number to use for the cost of energy for a HPWH. Especially in the context of this thread for using *surplus* solar for heating water, energy from solar is <$0.05/kWh surplus would be effectively free since it's wasted if it's not used. HPWH are what I refer to as 'discretionary' load. If you have a 80gal tank you use the water and it 'recharges' when it's convenient for the grid. CAISO is already tossing >200TWh/month and that's only going to get worse as solar penetration increases. Increased electrification is key and water heaters are the best place to start. The new Rheem HPWH even have a wifi connection so all that's needed to use them for grid balancing is software.

So if you were to compare PV + HPWH with a COP of 3 you're looking at (0.05)/(3) = $0.017/kWh

VS

1 therm is ~30kWh.

$1.51/therm => $0.05/kWh

And that cost of gas doesn't include the cost of gas service. I personally don't understand the benefit of even having gas service in areas that don't regularly see temperatures <-20F. The first thing I did when I bought my house was get an electric stove, mini-splits and HPWH so I could disconnect my gas service. I produce >2x as much energy as I need from solar.... didn't make any sense to keep importing fools fuel. It makes even less sense to throw out TWhs of solar energy because it has no place to go.... then use fools fuel to heat water.....

[bcroe]

As for improvements in air-air HP systems, I'm pretty up to date on what's happened and
what may be on the horizon. Seems to me most of the improvements of the last 20-30 yrs. or so come from:
- Improvements in the Thermodynamic properties of the working fluids.
- More flexibility in the operating speeds and working fluid handling (for example, multispeed compressors and air handlers).
- Some improvements in mechanical operation (scroll compressors, etc.)
- And what may be the biggest improvement: Massaging the hype and sales pitch to make it easier for users to think the actual improvements produce more efficient operation with the implied or vaguely promised associated improvements in unit cost effectiveness. All the bogus crap about EER/SEER comes to mind. Seems to me that's a better tool for B.S'ing users than it is for users who don't understand any of it, but rely on the hype. Seems to me a lot more objective reality would be conveyed by simply sticking to COP. SEER's of 20 - 30 are mostly B.S. anyway as is the whole slippery EER/SEER scheme.

One thing that doesn't seem to have improved and that mfgs. don't seem to talk about much, and as best as I can tell isn't accounted for in anywhere is the method and energy used to get the frost off a HP heat exchanger when the coil temp. drops below the frost/dew point temp. in the winter.

FWIW and IMO only, most of those improvements tend to benefit operation in cooling mode. The lower COP's as f(outdoor amb. temp.) in heating mode make it more difficult to obtain the same cost effectiveness as is possible with nat. gas for heating. The same applies for HPWR.

For my 2014 heat pump (SEER 14.5) in heat mode, the heat to defrost the outside coil came
from temporarily reversing to use house heat. This was quite obvious as I heard a CLANK
from the reversing valve, and the frost melted with the outside fan off as I got cold air. I have
no doubt my 2018 mini splits (SEER 33) do the same, but it is not nearly as obvious. They are
silent,and I think only need half as much house heat to defrost.

These units claim more BTU output in heating mode, than in cooling mode. That I would expect,
because the heat output always contains the operational power as well as the pumped heat.

Sounds like nwdiver has an HVAC and solar setup similar to mine.

Bruce Roe

[NochiLife]

Getting back to actual PV solar water heating, it works for me. This is a chart of about 11 days of water heating taking showers, dishwasher, etc. Temps drop down overnight to about 107 and this thermostat (likely because of litigation threats) can only go up to 127F even though the data sheet says 150. It normally shuts off about 10am. It doesn't prove anything except it meets my needs. The one thing that really shocked me is that it starts heating about 7am. And this energy is from the same panels that recharge my battery. Data was taken by an Elitech RC-5 which for a cheap data logger packs a lot of features and every data point on the chart can be accessed just moving the mouse. I think I sampled every 5 minutes. morning_water.jpg

Based on your earlier post, should I safely assume that your solar array produces more power than the charge controllers can take? (If the water heater is producing heat at 7 AM...does that mean that there is excess power available then...?)

[J.P.M.]

These units claim more BTU output in heating mode, than in cooling mode. Bruce Roe

That is an honest claim as can be verified by consulting any text on Thermodynamics. Or again, and if interested, see Wikipedia for more info. Readers' Digest version: Even after accounting for the parasitic energy needed for coil defrost for units using atmospheric air as a heat source (or think of ground source heat pumps that won't need coil defrost), the COP in refrigeration mode is always, at least theoretically, numerically 1 unit less than the COP of the same unit operating in heating (heat pump) mode. As you note, in heating mode the input work becomes part of the energy sought. It's just the way the Thermodynamics works out.

So, a unit with a theoretical cooling (A/C) COP of 3.0 will have a theoretical heating (heat pump) COP of 4.0. Real world parasitic losses, because they becomes part of the energy that costs in either mode will lower that 1.0 value, but coil heating parasitic losses/loads will only affect heat pump mode, and won't affect cooling mode as much, if at all. Those coil heating loads can be hard to estimate as they are f(coil temp. - dew point temp., atmos. dry bulb temp., coil design, other local conditions). Another reason why SEER is a bogus concept.

I haven't yet and don't plan to run the numbers or do the research so I won't challenge it too hard, but an SEER of 33 works out very roughly to a COP in heat pump mode of something like (33*0.875)/3.412 = 8.46. Based on what I think I know, I find that number hard to believe from air source heat pumps, even with heat source atmospheric air temps. high enough to not need any space heating in the first place. Too many numbers don't add up for me. Example: If it takes 85,000,000/0.9 = ~ 94,000,000 BTU (~ 27,700 kWh) to heat a dwelling for a year in a cold climate with a modern condensing gas furnace, that would mean 33 SEER heat pump(s) would reduce that energy demand to 3,300 kWh/yr. I don't see it.

Respectfully,

[bcroe]

Another reason why SEER is a bogus concept.

I haven't yet and don't plan to run the numbers or do the research so I won't challenge it too hard, but an SEER of 33 works out very roughly to a COP in heat pump mode of something like (33*0.875)/3.412 = 8.46. Based on what I think I know, I find that number hard to believe from air source heat pumps, even with heat source atmospheric air temps. high enough to not need any space heating in the first place. Too many numbers don't add up for me. Example: If it takes 85,000,000/0.9 = ~ 94,000,000 BTU (~ 27,700 kWh) to heat a dwelling for a year in a cold climate with a modern condensing gas furnace, that would mean 33 SEER heat pump(s) would reduce that energy demand to 3,300 kWh/yr. I don't see it.

Respectfully,

Agreed, SEER is not a reliable figure. It is all I have on these for now. I only take it to mean
that 33 SEER are more energy efficient than 25 SEER, but I could be convinced that even
that is a distortion.

That said, there are a lot of other factors favoring my later heat pumps. One is an ability to be
effective some 3 dozen degrees lower outside air, than my earlier unit. I am sure the actual
COP is riding up and down against air temp, but that mode still has a huge advantage over
switching to resistance heat, COP of 1. Bruce Roe

[J.P.M.]

Agreed, SEER is not a reliable figure. It is all I have on these for now. I only take it to mean
that 33 SEER are more energy efficient than 25 SEER, but I could be convinced that even
that is a distortion.

That said, there are a lot of other factors favoring my later heat pumps. One is an ability to be
effective some 3 dozen degrees lower outside air, than my earlier unit. I am sure the actual
COP is riding up and down against air temp, but that mode still has a huge advantage over
switching to resistance heat, COP of 1. Bruce Roe

Thank you. Understood.

I'd still think a comparison of the LCOE of any method, including various heat pumps of any sort against any and all space heating methods that are potentially or preferentially possible would be a good idea including, potentially and depending on the application, the energy sources of nat. gas, wood, passive solar, active thermal solar and PV, and any other source available and under consideration either by preference or limitation. I appreciate that the economics of energy use is not the only consideration, but it's often the deciding factor - (but probably not for you and me and some other eccentrics - I'm damn sure I can't justify the solar stuff on my roof based on economics alone and I bet you are in a similar situation with your stuff) - But for most folks who consider PV to cut or eliminate electric bills to limit the financial analysis with respect to HVAC methods to a choice among one source - as in the case of space heating limiting it to heat pump choices seems, to me at least, limited in scope if optimum cost effectiveness is going to be the deciding factor.

Respectfully.

[J.P.M.]

$0.35/kWh is at best a disingenuous number to use for the cost of energy for a HPWH. Especially in the context of this thread for using *surplus* solar for heating water, energy from solar is &amp;lt;$0.05/kWh surplus would be effectively free since it's wasted if it's not used. HPWH are what I refer to as 'discretionary' load. If you have a 80gal tank you use the water and it 'recharges' when it's convenient for the grid. CAISO is already tossing &amp;gt;200TWh/month and that's only going to get worse as solar penetration increases. Increased electrification is key and water heaters are the best place to start. The new Rheem HPWH even have a wifi connection so all that's needed to use them for grid balancing is software.

So if you were to compare PV + HPWH with a COP of 3 you're looking at (0.05)/(3) = $0.017/kWh

VS

1 therm is ~30kWh.

$1.51/therm =&amp;gt; $0.05/kWh

And that cost of gas doesn't include the cost of gas service. I personally don't understand the benefit of even having gas service in areas that don't regularly see temperatures &amp;lt;-20F. The first thing I did when I bought my house was get an electric stove, mini-splits and HPWH so I could disconnect my gas service. I produce &amp;gt;2x as much energy as I need from solar.... didn't make any sense to keep importing fools fuel. It makes even less sense to throw out TWhs of solar energy because it has no place to go.... then use fools fuel to heat water.....

For starters, I'll stick to how you incorrectly handle the savings by using POCO reimbursement rates and not discuss the relative merits/drawbacks of HPWH vs. nat. gas water heating.

I don't believe my use of the super off peak rate of $0.35/kWh from my POCO for power is disingenuous as much as it was quite conservative. It's the lowest rate available for a tariff most customers of SDG & E are on now or soon will be on using a T.O.U. tariff. Using the lowest rate assumes all water heating is done only at times of super off peak rates. While that's certainly possible, my guess is many or most residential users have no concept of what I just wrote, much less do it. If I had used some blend of hourly per kWh rates rates based on a more likely usage pattern, the cost/kWh would have been closer to something like ~ $0.40/kWh or so, making the more likely HPWH scenario I used even less cost effective compared to nat. gas.

Disingenuous: Not candid or sincere, typically by pretending to know less than one really does.

To my memory, your implication that I pretend to know less than I do will be the first time anyone's accused me of that behavior. Often, I'm considered at least a condescending prick, and a cross between a computer and a prostitute, that is, an f-bombing know-it-all, but that's off topic.

As I've not checked it, I accept your 200 TWH/mo. as excess, and even though I may not understand your reference and assume it has something to do with CA statewide excess PV production, but I'm not sure how much that has to do directly with the economics of residential excess energy production.

And, even though it's also probably off topic, I also disagree that increased electrification is always the best way to go. FWIW, the way I learned it many years ago, the better way to meet an energy need is to match the quality of the source of energy to the quality of the energy required for the task. Don't cut butter with a chain saw. Electricity is a very low entropy (very high quality) form of energy. That low(er) entropy and high quality gives it a lot of versatility. Example: You can heat water with electricity, but try directly powering the screen your looking at with nat. gas or wood. The nat. gas can be used to generate electricity at something like 40 % production eff. at a central generating station with the gained advantage of more flexibility of the product (the electricity put on the grid) to do more tasks than the nat. gas by itself. To then use that electricity to do something the gas could do directly (heating water) wastes that versatility and the money spent on the electricity generation.

Now, and more on the topic of your use of POCO reimbursement rates as a measure of cost effectiveness: Do and think as you want, NOMB or concern, but perhaps to help others not well versed in such subjects avoid thinking your logic is correct and to be followed, know that using the reimbursement rate of $0.05/kWh, or any reimbursement amount as a basis for costing as you have done is way off the mark and incorrect.

The correct number to use is the avoided cost of the electricity used to meet the duty required for the task. Reimbursement can improve the excess generation economics some, but usually only a little, and only fully when the reimbursement price/kWh == the cost of PV generation per kWh. That rarely happens. In the case of residential water heating using electricity, that cost will either be the cost of power from the grid, or the cost of PV generated electricity, or some blended mix of rates, not the reimbursement price paid for excess PV production.

To keep it simple, on topic and not muddy the waters, assume the cost per kWh of rooftop PV generated electricity is the same as the price paid for grid tied electricity by whatever costing method the party paying for the power chooses. Also assume a HPWH replaces an existing electric resistance tank type water heater, and the existing PV system exactly offsets electricity usage before changeout to a HPWR. Also assume that all other electricity uses remain the same, making the HPWR usage vs. the electric resistance heating the only usage change. The PV system also remains unchanged and is sized to meet the entire pre-changeout annual electrical load. I appreciate the system annual output will change, but for the sake of simplicity, assume it's constant year to year. If that's not acceptable use total modeled PV production for the system lifecycle divided by the number of years used in the lifecycle.

The HPWH will use less electricity to meet the water heating task than the old tank type resistance heater. So, for this example and conditions, the annual household electricity use will drop by an equal amount. That will result in an equal electrical bill (and because the PV is sized to replace the entire electrical load == $0 or some annual min. payment), but excess energy production will result in an excess production payment to be made at some low rate such as the $0.05/kWh you use. Keep the overgeneration payment separate from the electric bill for now.

Next, consider the cost per kWh of PV generated electricity. Say a system of 10 STC kW is in place, cost the owner $21,000 after tax credits and generates enough electricity to exactly meet the annual pre HPWH load as in the example. For reference, say that annual generation and annual use before the HPWH changeout is 16,000 kWh/yr.

The PV system was paid for in cash up front. It's a sunk cost and so cannot be recovered. The fact that it's a sunk cost is one key here. You won't get any of that money back by energy improvements.

The HPWH changeout will do nothing to reduce the annual cost of owning the PV system. Since PV generation already offsets 100 % of usage before the changeout, what the changeout will do is reduce the amount of the PV produced electricity that goes to offset (now reduced) usage that formerly carried a production value of the retail electricity rate and now has a value of $ 0.05/kWh. After the changeout, the resulting excess production is only gaining $0.05/kWh. Before the changeout, that production was valued at the much higher retail rate.

Look at it another way and still using my simplified example with some numbers using what I consider to be my genuous, conservative but still traceable and actual $0.35/kWh rate.

From the example, 16,000 kWh/yr * $0.35/kWh = $5,600/yr. annual production value and annual electricity cost with the result of a zeroed out bill. Now, reduce that annual load by, say, 1,000 kWh/yr. as a result of the changeout from electric resistance heating to a HPWH. The bill is still zero or some min. charge, but the cost of the PV system is unchanged (remember the sunk cost ?). So, for one thing, the benefit from the reduced usage is only the excess generation credit, but the $5,600- $300 = $5,300 net annual cost will lower the per kWh cost effectiveness of what's now effectively an oversized system. What's happened by the changeout is that, in effect, the system revenue has been lowered by ($0.35-$0.05) * 1,000 kWh/yr. = $300/yr. as a result of the underutilization of the now oversized system.

As an aside, that's also a way to see the wisdom of making as many lifestyle changes and energy improvements as possible before considering PV rather than after.

Although it may seem a bit like closing the barn door after the horse has gotten out, and without writing or implying anything about the economics of HPWH, and using the example's 1,600 kWh/hr. PV production per installed STC kW and $3.00 * 0.7 net installed cost/STC W, the system sunk cost could have been reduced by (1-(15,000/16,000))*$21,000 = $1,312 if the HPWH changeout had been done before PV was considered and the annual electric bill would still have been zeroed out. Perhaps coincidentally, I'd point out that $1,300 or so is in the neighborhood of the current material cost of a HPWH.

An analogy using fuel costs associated with driving an ICE vehicle: Say I buy a new car and plan on keeping and maintaining it for at least 10 years. I drive 30,000 miles/yr and my shiny new ICE vehicle gets 30 MPG. Also say gasoline currently cost $3.00/gal. I therefore will burn 1,000 gal. of gas/yr. at a cost of $3,000/yr. Now, my friendly gas station comes to me with a deal: For the up front consideration of $30,000, they will provide gasoline to me at the rate of 1,000 gal./yr. for the next 10 years. Anything > than 1,000 gal./yr., I pay the going rate. Anything less than 1,000 gal./yr., they'll credit me at the rate of $0.05/gal. with no yr./yr. carryover. I take that bet because, among other reasons, I feel gas prices will only go up. Now, about 6 months into my 10 year deal, I see an ad for a super duper add on to ICE vehicles like mine that will guarantee to increase my gas mileage by 10%. I check it out and it's a true and honest claim. I continue to drive 30,000 miles/yr. Q: how much do I generate in fuel savings per year by getting the add on ? Answer: Since I've already paid $30 large my savings will be Zero $$. Cost: The cost of the add on device.

Point for those considering PV: Do your homework before getting PV and keep sight of the goal of lowering the electric bill as opposed to getting PV for it's own sake or an end in itself, and get the costing right.

Take what you want of the above. Scrap the rest.