KB's Design

[kb58]

Or what passes for one.

I already have an intro thread but not a walk-through of my planned build, so here goes.

Currently I'm finding it surprisingly involved to determine what size system is needed, so while I flounder around with that, I at least did a site survey. Our house is really small and not oriented well for panels, and due the topology of the yard, a ground mount works out better. We have a terraced rear bank that runs NE-to-SW at 220 degrees, with significant shading toward the east which I can't do anything about - see the attached "Solar Shades" scan of the site. The terraced rear bank has a ~10-ft wide flat "step" which contain a concrete drain swale with the likely PV area being approximately 11'-wide x 66'-long. I want the panels as low profile as possible because whatever's up on the bank is very visible not only from the house but the street as well. I'd like to orient them in landscape format, 1 tall x 2 wide, spaced in rows to avoid self-shading. The front edge would be just high enough to avoid small weeds, tipped at around 25 degrees and facing 220 degrees. You can see from the attached that coincidently, that's probably about perfect. The 10-11' long assemblies will straddle the concrete drains, which isn't a usable space anyway, so that works out well. The maximum height should be around 24" or so which minimizes the eyesore and the wrong kind of attention. (As an aside, pole mount setups are doubly bad, both in terms of visibility and because the terrible clay/sandstone soil making boring deep holes nearly impossible. The rear bank also has no access for heavy machinery.)

We've had one quote done ($18,000) which recommended a 5.6KW setup. The thing is, shortly before the quote was done we redid our koi pond, increasing power usage somewhat. As a result, the quote's based on just a few months of increased usage, and they estimated the yearly monthly average to be around 700KWh. I'm pretty sure that's going to prove a bit low but we'll see as the months progress. As I mentioned in the intro, me being an EE, I gradually decided to do the installation myself (one of my hobbies is building cars from scratch, the latest being a mid-engine two-seater resembling a Lotus Seven - Goggle "Midlana" to go there, but I digress. Other future additions to the electrical system may include a EV and/or a battery depending on incentives, but I'm getting ahead of myself.

Anyhow, due to the very helpful people I've already met here I know I have my work cut out for me just determining panel count. I guess the first thing is to get a better estimate on the average daily or monthly usage. I should be able to extrapolate the usage for the months we have to get something fairly accurate.

Other stuff... we use SDG&E, so TOU 2.0 is looming for next year and will be defined as 4-9pm. While this makes it more difficult to fill that time slot with power, as you can see from the site plot, it's pretty clear off toward the west. If I choose, the panels could be angled even straight west if that ends up being economically the best direction.

Enough typing for now, more later.

[max2k]

Or what passes for one.
...
We've had one quote done ($18,000) which recommended a 5.6KW setup. The thing is, shortly before the quote was done we redid our koi pond, increasing power usage somewhat. As a result, the quote's based on just a few months of increased usage, and they estimated the yearly monthly average to be around 700KWh. I'm pretty sure that's going to prove a bit low but we'll see as the months progress. ...

in terms of required size it helps if you could analyze/understand where those 700kWh go in case you have some 'sticking out' load(s) which can be easily accounted for. Hunting down to the last watt could be too time consuming. Have you upgraded your incandescent bulbs to LED? Those bulbs were responsible for at least 25% of my original bill. Old computers running 24x7 might also worth a look. There's a device called Kill-a-Watt which allows you to measure how much particular load consumes over time. I didn't use it as I had easily identifiable offenders. Please keep in mind people commonly increase their consumption after installing solar by at least 10% mostly by setting AC lower in summer.

[kb58]

Completely agree on conservation; we're nearly all LED, especially the lights used the most. What takes the most power though is the pond system, nearly 800W 24/7. With TOU 2.0 coming I'll shut off the non-critical loads during high-time, which should save around $17/month, not a lot but it adds up. Agreed on obvious outliers, and the main pond pump is it, at around 370W round the clock. That, and the 120W UV can be shut off during high-time, but it's not the sort of thing which can be left off for long, but that'll be looked into.

[J.P.M.]

On the new NEM/T.O.U. tariff scheme: You will probably be on tariff DR- SES. I'd encourage you to continue to read up and understand what that means to you in all its muddled glory.

One perhaps neutral - to - not so bad consequence of DR-SES is that, because its per kWh rates are not determined, nor are they f(usage), for sizing and revenue estimates, any model's output will, for any set of hourly output in kWh from, say, PVWatts for example, or any other model that produces system hourly output, the value of a system's output as modeled will have a constant $$ value to offset a bill under DR-SES (at least as long as those rates and times hold), provided the system is sized for < or = to the annual kWh usage.

In other words, for as long as the tariff rates and times for DR-SES stay the same, and provided the system offsets <100 % of usage, the $ value of the system's output as modeled that will be used to offset a bill will have a constant value regardless of the amount of usage or the times of that usage.

That can make system sizing and array orientation a lot more straightforward, and help highlight the consequences of time shifting usage (or not time shifting usage), for both good and bad effects .

An example: Say your annual usage is 10,000 kWh, and never mind the time pattern of hourly usage for now. Also say that you use a spreadsheet, DR-SES tariff rate and time knowledge, and an hourly PVWatts output, and determined that a 1 kW array orientated at 30 deg. tilt and 225 deg. azimuth will, as a long term average, have a pretty good chance of producing about 1,690 kWh/yr, and, under DR-SES and estimated rates from SDG & E for the new tariff, offset (or feed to the grid under NEM) $390 of your electric bill, regardless of what the $$ amount of that bill for usage may be, or how much electricity is actually used.

Now, to offset 100 % of the original assumption of 10,000 kWh/yr. usage, (not necessarily 100% of the bill just yet), using the above model/spreadsheet numbers you'll need a system of ~~ 10,000/1,690 = 5.91 = 6 kW. However, and this is where things get a bit muddy, if you use all your electricity at off peak rates, that is, zero usage during peak pricing times (which is unlikely, but stick with me for minute here), your average per kWh cost will be something like, and very approximately, $0.22/kWh. That will make your annual bill for the 10,000 kWh = ($0.22)/10,000 = $2,200. Here's the trick: Since the value of the electricity that the PV system produces stays constant at $390/STC kW per yr. regardless of how much electricity you use, or when you use it (provided, for this example, that it is not used at peak pricing times), your bill, by avoiding peak times is $2,200. That means a system size = ($2,200/yr.)/($390/yr. per kw) = 5.64 kW, or slightly smaller will offset your entire bill for the same 100 % offset of the bill.
Note that the system output will still be 1,690 kWh/STC kW per yr.or 5.64 * 1,690 kWh/STC kW per yr.= 9,532 kWh/yr., or less than annual usage and still offset all the bill, less $120/yr. min. charge accrued at $10/month.

Or, another example going the other way on times of usage: Suppose all of your use is in peak pricing time, another equally unlikely scenario, but simple to see the purpose of the example. Under this scenario, say the per kWh rate for use during peak times is $0.46/kWh. That means your annual bill will be ($0.46/kWh) * 10,000 kWh/yr. = $4,600/yr.

To offset that $4,600/yr. bill, and because system revenue to offset a bill is still effectively fixed at $390/STC kW under DR-SES, a system size of ($4,600/yr.)/(390 $/yr per STC kW) = 11.79 kW will be needed to offset that entire $ value of the bill.
Note that system output will still be 1,690 kWh/STC kW per yr., or 11.79 * 1,690 kWh/STC kW per yr. = 19,908 kWh/yr.

Now, since that size system will produce about 2X as much output as usage, there will be excess generation for which you will be paid, at current rates, something like the "giant" sum of $0.027/kWh or ~ = (19,908 - 10,000 kWh) * $0.027/kWh = $ 267.51/yr. That's one way the constant revenue part of what I'm writing about fails and why system output needs to be <100% of usage for the constant revenue assumption to stay valid.

Kind of an off topic sidebar here: Note that shifting 1,690 kWh/yr. from peak to off peak time can reduce system size by ~ 1 STC kW , and reduce the annual bill to $ 4,600 - [(1,690 kWh/yr output per STC kW)* ($0.46-$0.22)] = $4,600 - $405.60 = $4,194.40 which, at $390/STC kW revenue will reduce actual system size to $4,194/$390 = 10.75 kW or an actual size reduction of 11.79 kW - 10.75 = 1.04 kW. The difference, 1.04 - 1.0 is in the reduced excess production payment and a consequence of the relaxation of the constant $ value of production from excess production requirement.
Point is, at, say, in addition to ($3.25 * .7)/STC Watt) *1,000 W = $ 2,275 initial savings/cost reductions that are possible by some time shifting, an additional $405.60 less some small decrease in excess production revenue from SDG & E, all savings possible from some time shifting. Again, the example is not realistic, but made overly simple to help explain how time shifting can be used to perhaps game the T.O.U. system and minimize, system size while enabling some perhaps easier way(s) to help determine tradeoffs between likely (in)conveniences of timeshifting scenarios vs. both initial and ongoing savings from time shifting strategies.

Bottom line: Under DR-SES which all PV users under SDG & E billing will be under, at least by default, for any system orientation, modeled system output will have a pretty much constant value to offset any bill, pretty much regardless of usage amount or times of that usage, provided system output is less than the annual usage. There are a couple of considerations of min. annual bill and NBC (non bypassable charges) to consider as well, but they will not change the value of modeled system output. Only changes to the DR-SES tariff will do that, and only because that tariff's rates, at least until the DR-SES rates and times change, are not linked to the amount of customer usage. NOTE: This is NOT true for all other tariffs and rates, at least for SDG & E. I believe the same can be said for other CA I.O.U.'s, but I'm not as familiar with those as I am with SGG & E's tariff situation.

[kb58]

Thank you for the very detailed response, I'll review it in detail later.

Somewhat related, because of this likely ruling about import PV prices, should I consider the kneejerk reaction of immediately ordering panels now instead of when I actually need them? I don't know enough about the market forces to know if the potential increases are real, or to what degree, and how much of it is media drama. At the moment I'm considering Jinko panels with the Maxim optimizers built it, which avoids the RFI generated by Enphase microconverters and SolarEdge optimizers, but am not certain I want to go that way yet (in addition to them being poly instead of mono). I'm not normally one to jump in early, especially to something so expensive, before I've done a detail analysis of what I want. That said, if PV prices are expected to "double" (like anyone really knows for sure), that's a big component in deciding when to pull the trigger.

[J.P.M.]

Thank you for the very detailed response, I'll review it in detail later.

Somewhat related, because of this likely ruling about import PV prices, should I consider the kneejerk reaction of immediately ordering panels now instead of when I actually need them? I don't know enough about the market forces to know if the potential increases are real, or to what degree, and how much of it is media drama. At the moment I'm considering Jinko panels with the Maxim optimizers built it, which avoids the RFI generated by Enphase microconverters and SolarEdge optimizers, but am not certain I want to go that way yet (in addition to them being poly instead of mono).

Upside: You'll perhaps garner a lower price.

Downside:
- Without proper array sizing, you may wind up over/underbuying.
- If U.S. protectionism does occur, it may make what you buy an orphan as panel mfgs. fall by the wayside/drop from the U.S. market.
- DR-SES prices are not yet firm. Close maybe, but whatever they wind up being, the cost effectiveness of PV under DER-SES, or the T.O.U. EV rates will be a whole lot less cost effective, probably by very ~~ 22 - 25 % or so than under the old rates and T.O.U. times. That will, if free market forces prevail, tend to decrease PV prices, otherwise cost effectivness, ROI and payback times will be adversely affected. An example: under the old DR-SES pricing and the more favorable to PV times that prevailed, the average value/kWh of PV generated electricity is currently/was ~ $0.31/kWh. Under the new rates/times, that average value to offset a bill will drop to ~~ $0.23/kWh. To maintain the same cost effectiveness, system prices will needs to drop approx. the same percentage.

If new protectionist tariffs raise prices like tides lift all boats, system prices will go up making already tariff induced cost effectiveness lower cost effectiveness even lower.

NOMB, but I'd get my tariff information, usage and sizing/ststem design information down pretty hard, get preliminary design done, call a couple of responsible electrical contractors who sell PV and get some quotes (while being very cautious of their costing information due to the newer, less favorable tariff situation that they will probably try to ignore or downplay for obvious reasons). In the mean time, if DIY is still, or gets on the radar, shop prices and get ready to move, or not.

Remember, in the end, PV, while it's a darling of the follow the lemming crowd, and perhaps not a bad choice (if done properly), is still a choice that may or may not make financial sense for some, maybe even most situations. I've seen a lot more bad choices and crappy installs due to solar ignorance than I have seen cost effective, fit for purpose residential PV systems.

Knowledge is power. The more you have of the first, the better use of the second.

[sensij]

An example: under the old DR-SES pricing and the more favorable to PV times that prevailed, the average value/kWh of PV generated electricity is currently/was ~ $0.31/kWh. Under the new rates/times, that average value to offset a bill will drop to ~~ $0.23/kWh. To maintain the same cost effectiveness, system prices will needs to drop approx. the same percentage.

It is worth pointing out that switching from DR to DR-SES will impact the *cost* side of the equation as well. For some, the same consumption would cost less under TOU than under the tiered plan, which implicitly adds to benefit of the PV system (imagine installing a 0 W system, just to get access to the tariff). For others, the opposite is true. Revenue calculations are very good at optimizing the orientation of the array, but are not enough of the story to deeply evaluate cost effectiveness.

Another point in favor of expediting the installation is that if you can get PTO before Dec 1, the new TOU hours won't apply for 5 years. Taking extra time to get the array size perfect and missing that deadline may not be the most cost effective plan.

[kb58]

Yeah I kind of wish I'd started this about 5 years ago.

[J.P.M.]

Yeah I kind of wish I'd started this about 5 years ago.

If, 5 yrs. ago, I knew how things would be 5 yrs. hence, that is, now, or the near future from now, I'd perhaps not have bothered with PV back then. Hindsight is always 20/20. I did well on it for other reasons, but in & of itself, it looks to me that from a cost effective standpoint, PV will not be as good a deal doing as it has been in the past.

[kb58]

Well that's certainly a possible outcome of the research, that it ends up as, "I came, I saw, and decided it's not for me."

[philips]

What takes the most power though is the pond system, nearly 800W 24/7.

Out of curiosity, what kind of flow rate do you need to run for your pond? Any static head?

I'm ignorant on Koi pond pumping requirements and filtration, but just for another data point, I run a variable speed pump on a pool at 15 GPM / 900GPH (pump setpoint) , 0 static head, ~15' dynamic head when the filter is clean (based on pump curve) and I use a little less than 200W. This rises as the filter loads up to about 350W, and at that point I back flush the thing.

[J.P.M.]

Well that's certainly a possible outcome of the research, that it ends up as, "I came, I saw, and decided it's not for me."

Deciding to not do something is also a viable option for most situations. Action is not necessarily a mandate, particularly if other option involving no action make better/more sense. Trick is to avoid dichotomous, all/nothing thinking. Partial bill offset w/PV after conservation/lifestyle mods. and time shifting can usually bring better cost effectiveness than throwing PV at a large, self inflicted electric bill.

[kb58]

Out of curiosity, what kind of flow rate do you need to run for your pond? Any static head?

I'm ignorant on Koi pond pumping requirements and filtration, but just for another data point, I run a variable speed pump on a pool at 15 GPM / 900GPH (pump setpoint) , 0 static head, ~15' dynamic head when the filter is clean (based on pump curve) and I use a little less than 200W. This rises as the filter loads up to about 350W, and at that point I back flush the thing.

6000 gallon, running about 5000gph through the bio filter, and 3000-4000gph through the sand filters. FWIW, there is a kick-butt variable-speed pump that ponders love which is very efficient called "Flowfriend". The catch is that it's $2700 and - at least for my use - would literally take decades to pay off the price premium.

Static head is zero and dynamic head is around 5-8 ft.

[philips]

6000 gallon, running about 5000gph through the bio filter, and 3000-4000gph through the sand filters. FWIW, there is a kick-butt variable-speed pump that ponders love which is very efficient called "Flowfriend". The catch is that it's $2700 and - at least for my use - would literally take decades to pay off the price premium.

Static head is zero and dynamic head is around 5-8 ft.

Thank you. Yeah, at $2700 it would likely die before it paid for itself.

[kb58]

Deciding to not do something is also a viable option for most situations. Action is not necessarily a mandate, particularly if other option involving no action make better/more sense. Trick is to avoid dichotomous, all/nothing thinking. Partial bill offset w/PV after conservation/lifestyle mods. and time shifting can usually bring better cost effectiveness than throwing PV at a large, self inflicted electric bill.

Partial bill offset is where things are probably headed, enough to get down to T1