Comparing bids - Production Same -Energy differs signficantly?

[rjblead]

Need a little help. I've got two vendors giving me what feels like conflicting info and I need some informed guidance like I find on this forum.

I'm trying to pick between two bids for solar on my new (almost done being built) home in sunny California. Given that it's new, I don't have a track record of consumption, but based on usage at my old home and some adjustment for other upcoming factors (electric vehicle, pool, etc.) I' ve asked them to assume I will use around 19,500 kwh annually. I talked to 5 vendors and ended up with these two bidders - one for SunPower and one for LG. Here's the rub of where I'm stuck:

Vendor A (Sunpower) - 11.52 Kw (DC) system with 32- X22, 360 Watt panels using their integrated micro inverter. They estimate this will produce approximately 19,167 kWh annually. Their cost (pre any rebates) is $45,800

Vendor B (LG) - 11.52 Kw (DC) system with 32-Neon R, 360 Watt panels using EnPhase IQ 7+ micro inverter. THey estimate this will produce approximately 15,576 kWh annually. Their cost (per any rebate) is $35,300.

Notice the overall power (11.52) is the same on both bids, but there's a big difference 3,500 kWh in estimated energy output. More importanly, there's a huge difference in cost! Is there some reason the Sunpower should put out a lot more energy? I assumed the power would dictate a roughly common output other than a slight difference in temperature coefficient. Also, the LG will likely have a slight bit more degradation down the road.

I asked both vendors to help me account for the difference. The Sunpower guy said the difference was because the Sunpower panel is much more efficient (I'm not sure why that enters into it) and that the SunPower inverter is much more adept. He says the LG guys is also likely underestimating output. The LG guy says he uses a fairly conservative, industry accepted estimating model (can be found at www.csi-epbb.com).

So what do I make of all this. If the LG guy is right about energy output and I go that route, then I save a bunch of money up front but still have a huge Electric bill each month (we're on Pacific Gas & Electric which just charge outrageously even at their lowest tier).

Is it right that they should both create about the same amount of energy day 1, regardless of what model you choose to estimate it? If so, then I think I'd be smart to go with the LG bid.

Thanks for the help

[faraway]

Not an expert, but the output will be around the same if the panels are positioned the same way.

[J.P.M.]

Do this:

1.) Take neither one's estimate as gospel. Scrap any idea of buying Sunpower if cost effectiveness is of any concern to you.
2.) Download and run PVWatts after you read all the help screens a couple of times. Get the orientations and tilts as close as possible and use 10 % system losses rather than the 14 % default the model starts with.
3.) If you do the input correctly and are somewhere near a true south orientation or so and maybe a 20 deg. tilt, an 11.52 STC kW system will produce something like ~ 19,000 to 20,000 kWh/yr.
4.) Know that either panel will perform about like the other one, +/- some small tolerance, and that you or anyone else won't be able to measure the difference.
5.) Disregard panel degradation. It's a small matter (much smaller than the price difference) and you'll never notice it, much less be able to ever quantify it. It's simply more Sunpower B.S/hype.
5.) While you're at it, spend 20 bucks or so and get a copy of :"Solar Power Your Home for Dummies" at bookstores/Amazon. You need an education. 20 bucks well spent.



I live in 92026 and watch my system like a hark as well as monitoring about 8 - 10 other systems in my HOA, 3 of which are Sunpower (like mine) and 3 others being LG. The rest are SolarWorld, Canadian Solar, etc. All the systems, once adjusted for orientation, produce about the same output. My system (tilt 19.5 deg., azimuth 195.7 deg., 5.232 STC kW) has produced 44,410 kWh over the last 1,774 days == >> 9,137kWh/365 days before shading loss for a 5.232 STC kW system, or ~ 1,746 kWh/yr. per STC kW. That output is after about a 4.5 % shade penalty from a SW tree that I don't own. Without the tree, my ave. output for the last close to 5 years be something like 1,830 kWh/yr. or so per installed STC kW.

PVWatts big brother,"SAM", estimates my unshaded output at about 1,809 kWh/STC installed kW. The other LG arrays perform similarly to mine as do the other Sunpower arrays and all the rest.

PVWatts output comes out about the same as SAM's if I use 9.6 % or so for a system loss factor.

Check PVOutput.org for arrays near you. Bet you'll find similar results in your area once adjusted for orientation.

Perhaps interestingly, "SAM" 's output for my location has LG with a slightly higher (on the order of 1% or so) annual output than the Sunpower units. I have no trouble believing that.

Bottom line for you: Run PVWatts and use the numbers you get for any quality panel for system design and sizing. The PVWatts output will be similar, one panel mfg. to the next. Prices in So. CA are running ~ $3.00 - $3.25/STC W for turnkey stuff, No. CA may be a bit more, maybe not. Check something called the CSI database for installs and prices paid near you. The Sunpower peddler probably got close to reality about 2 things: Right about vendors underestimating output, and right about his exorbitant price - Sunpower's often a buck/watt more up front for no real long term benefit other than bragging rights. The underestimation of output is common - that's done so vendors can take advantage of most users solar ignorance, screw them over and put more product on a roof than is needed. No peddler ever got fired for selling too much. There job is to put product on your property, as much of it and for as much $$ as possible. Their job has little if anything to do with saving you money on your electric bill.

BTW, there's nothing much special one panel to the next. They are now pretty much a commodity like water heaters, toasters or garage door openers. Sunpower and GG both make quality stuff. They are not the only ones. LG is more price competitive than Sunpower, buy far, but LG is not the only game in town worth considering. Spend as much time on vendor evaluation and selection as product evaluation. With all the components now becoming a commodity, vendor quality and integrity, which has always been paramount, is more important than ever, particularly now that the mfgs. and vendor shakeout is underway. Vendors who were around before PV and are more than a 1 trick pony have a better chance of surviving than the "Larry with a ladder" types. Think most bang for the long term buck, not necessarily low initial cost.

Lastly, get familiar with how utility rates and NEM rules are changing. Vendors often don't mention that due to rate and NEM restructuring in CA, PV is about 20-25 % less cost effective than it was before all this rate change business started a couple of years ago.

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

Good luck.

[JSchnee21]

Definitely go with LG or Panasonic, or a less expensive, less efficient panel (LG Neon2, Hanwha, Trina, Jinko, etc.) if you have more roof space available and can still reach your desired DC kW target. I'd scrap the Micros (Enphase and SunPower) and go with Solar Edge instead. If you have zero shading, some here would advocate a standard string inverter from Fronius or SMA -- But, I prefer the panel level monitoring and rapid shutdown -- which is likely required (rapid shutdown) in CA these days (or will be soon).

To be honest, there are many, many different pieces of software which will attempt to model annual kWh based on a variety of inputs. Modelling and simulation accuracy is highly dependent on the model and input data, including a variety of hyperlocal factors which won't be captured. You can try looking at PVoutput (dot org) to find similar systems in your local area for comparison -- but be advised the orientation, pitch, and shading will have a large/huge impact on annual production. The SunPower estimate seems rather high to me, but the reality is your annual production will likely be somewhere in between the two.

Since you're in CA, which seems to have crazy and over complicated rate plans for electricity, the bigger question in my mind would be how your net metering, time of day production, usage, and net billing works. It seems to be that in many parts of CA power companies (PoCo's) are making an effort to screw over residential customers in the way in which they use price discrimination to adversely bill and limit PV credits.

Before making such a large investment, I would look carefully at the rate plans your PoCo has available in comparison to your current and future time of day usage and consider whether a small system (or same sized system) with a storage component might be more helpful to time shift your self consumption. Granted storage is still very expensive and often not cost effective, but some PoCo's in CA seem to have promotions and rebates -- not to mention the crazy peak kWh billing rates they penalize you with in the afternoon / early evening.

[rjblead]

JPM and JSchnee - Thanks for the super helpful replies. I'll be picking the LG solution. I ran my numbers through PV Watts and got a calculation of the anticipated output at 18,548. I don't have the roof space to go with the less efficient models. Loved pouring over the CSI data for my area to compare deals others have been getting. Got out to PVOutput.org and I'll become a contributor there whan I get this system in place. I'll defintely check into the NEM rules as recommended. Again, sure appreciate the help!

[J.P.M.]

Definitely go with LG or Panasonic, or a less expensive, less efficient panel (LG Neon2, Hanwha, Trina, Jinko, etc.) if you have more roof space available and can still reach your desired DC kW target. I'd scrap the Micros (Enphase and SunPower) and go with Solar Edge instead. If you have zero shading, some here would advocate a standard string inverter from Fronius or SMA -- But, I prefer the panel level monitoring and rapid shutdown -- which is likely required (rapid shutdown) in CA these days (or will be soon).

To be honest, there are many, many different pieces of software which will attempt to model annual kWh based on a variety of inputs. Modelling and simulation accuracy is highly dependent on the model and input data, including a variety of hyperlocal factors which won't be captured. You can try looking at PVoutput (dot org) to find similar systems in your local area for comparison -- but be advised the orientation, pitch, and shading will have a large/huge impact on annual production. The SunPower estimate seems rather high to me, but the reality is your annual production will likely be somewhere in between the two.

Since you're in CA, which seems to have crazy and over complicated rate plans for electricity, the bigger question in my mind would be how your net metering, time of day production, usage, and net billing works. It seems to be that in many parts of CA power companies (PoCo's) are making an effort to screw over residential customers in the way in which they use price discrimination to adversely bill and limit PV credits.

Before making such a large investment, I would look carefully at the rate plans your PoCo has available in comparison to your current and future time of day usage and consider whether a small system (or same sized system) with a storage component might be more helpful to time shift your self consumption. Granted storage is still very expensive and often not cost effective, but some PoCo's in CA seem to have promotions and rebates -- not to mention the crazy peak kWh billing rates they penalize you with in the afternoon / early evening.

I agree with the idea of looking at less expensive panels although I'd avoid calling them "less efficient - using the term "less efficient" gives it a validity that is not warranted and gives those who use such B.S. terms a victory over common sense.

I too would scrap the micros altogether, but I tend to favor keeping systems as simple as possible and also as much of the electronics off the harsh environment of a roof as possible. To me, that simply means a string inverter in the garage. As for panel level monitoring, I've found that users bore of it after a couple of weeks or so and ignore it thus putting any advantage it may have, in effect, in a drawer that seldom gets opened. Besides, most users are clueless about what they're looking at and if they do any thing at all, usually call the installer and complain when panels' output varies by a couple of Wh or less. Putting my money where my mouth is, I have a simple string inverter and I get more than enough data out of it - more than I'll ever use or can make use of - and believe me, I use a lot of data. The inverter is in the garage, safe, cool and out of the elements. Close to 5 yrs. in and it hasn't so much as burped (touch wood).

The way I look at it, any shade will decrease a system's cost effectiveness. For system design, first get a location and orientation as shade free as possible. If some shade, see if/how much of any shade penalty can be mitigated by micros, etc. and balance that against the cost remembering that the goal is to pay as little over the long term to provide electricity to a residence. That may, BTW, mean that PV may n ot be the best wat to achieve that goal.

Micros or such things as the SolarEdge system or such can perhaps recover some or even most of the energy harvest (and some of the revenue) that would otherwise be lost by use using a string inverter in a shade situatipon, but energy that was not there to be harvested in the first place because of shade cannot be recovered by any system. A poor application because of shade cannot be made as efficient as a non shade application simply by not using a string inverter. An extreme and unrealistic example or two might be: An array in complete shade 24/365. Think: How much will a micro or a SolarEdge system improve performance ? Another example: An array is completely unshaded, 24/365. How much will a micro or SolarEdge system improve performance ? The reality of an application is usually somewhere between.

I'd suggest that micros/SolarEdge can improve a marginal shading situation, but a lot of shade will be difficult to make an application cost effective the way things are now and appear to be going. Seems to me there are a lot of uninformed people being told, or are under the impression that micros/solar edge systems can make poor, or most any applications adequately cost effective when the reality may be something less.

As I wrote, get familiar with how you are charged for power. Having done this, I'll say doing that correctly is a real big PITA - be under no illusions - and not made any easier by the POCO's seeming intransigence at helping to make the process any easier. But, and I say this just as adamantly, doing so will pay big dividends, particularly if done before you get any PV. Do this step right during system design and I bet any system you subsequently acquire will be better designed and also smaller and more cost effective.

I'd disagree with the idea that the POCO's are making an intentional and malicious effort to screw over their customers who have PV systems. Reason: There is simply no long term financial gain in screwing people who you sell things to for the sole reason of screwing them, if for no other reason than if you piss people off, they are not as easy to manipulate. It's just business. Nothing personal. Doing things that way just makes life and making a profit more difficult. Doing what's thought necessary on the way to maximizing profit may make for some unpleasantries, but they are more considered decisions and perhaps more accurately thought of as more of the order of collateral damage than some nefarious plot or personal vendetta against PV users. There's simply no gain in the POCOs getting personal about it. I'm no fan of POCOs, but they are doing no more than their mandate of making a profit and going all they can to maximize that profit for their stockholders. It's one aspect of capitalism. T.O.U. pricing and particularly the times of peak pricing might be thought of as the POCOs' way of getting what they need while not aggravating too many folks. If so, that method seems to be failing, at least when it comes to not aggravating customers. However, I still don't believe the POCO rate structures are designed to intentionally screw PV owners in some malicious "get even" way. To think that, IMO only, is a childlike and self centered outlook. Besides, most folks have the ultimate ability to control the size of their electric bills if they'd wake up and smell the coffee - USE LESS ELECTRICITY. Seems to me most folks would rather simply go on with their profligate use unabated, while doing nothing more than bitch, cry and moan about what it costs, and then claim the robber baron POCOs have nothing more than a woody for them. I'd suggest more than a few PV owners are acting a bit childish, looking for a free ride and crying when the rules change. To folks who think such things I'd say, grow up and deal with reality.

On modeling software, if you start out knowing and accepting the idea that solar modeling is far from as exact a science as most people would like to believe, and less so than most peddlers would want to cause potential PV users to infer when they (the peddlers) wave some technical and semi official looking document with "prediction" numbers on it under customers' noses, you'll probably avoid the dismay, disappointment and confusion when a time period's system output varies by, say, 10 to 20 % or so from what the model predicted, which will not be at all unusual. A correctly modeled system's output gets closer to actual output the longer the period of analysis and record. That means a period that's probably measured in years, not weeks or months. Also, keep the idea in mind that, while you can use a model as a prediction tool, and it's tempting to use them as such, their primary purpose is for system design. There is a lot of confusion generated by thinking, for example: "model X 'told' me I'd get so may kWh the first week in Aug." and then calling the model a POS, or thinking there's something wrong with their system when actual system output was, say, 30% lower than what the model's output was for the same (short) period. Similar to the difference between climate and weather, with climate being what you expect and weather being what you get, a correct model's output for last week is what you expected, last week's output is what you got. The two numbers may be close by coincidence, but not by design.

On system design and storage interface: At this time and until things shake out a bit both in terms of where NEM is headed and where residential storage is now and will be, for many reasons I'd wait on adding storage (not the least of which reason being that early adopters wind up being beta testers and usually get systems with not all the bugs shaken out), but plan/design for a system that may be able to have storage "bolted on" or "wired into" it when the situation becomes more clear and storage becomes cost effective and standardized. SWAG on when ? Maybe 5 years down the road, but my crystal ball is still in the shop at this time, so my guess is no better/worse than the next guess.

I'd suggest one of the parameters affecting when and what size a storage system will make financial sense will be just when and in what ways peak T.O.U. pricing hours will evolve/change. An unlikely but theoretically possible scenario for thought only: Suppose EV's REALLY take off to the point where nite time home EV charging shifts the peak grid system demand way over to 0100 - 0600 hrs., and that change forces peak pricing to shift to those times. That could and probably would cause things to change in many ways. Overall, seems to me, at this time anyway, that the energy storage that will best allow avoidance of as much use of grid supplied power during peak pricing times as possible will likely be the most cost effective. But, I'd suggest, still nowhere near as cost effective as first, not using the power at all, and second, time shifting as much of that reduced load away from peak pricing as possible, both of which will have the added benefit of allowing a smaller size storage to be used (and spending less $$ to acquire it).


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