Sun hours

[Ampster]

Thanks for demonstrating that outside the box solutions can work and accomplish the owners goals.

[bcroe]

Hey design reviews are good and educational. I had a career of them with reviewers a lot tougher
than you guys. He whose design is being reviewed needs to see it as an oportunity to improve, it
certainly helped me.

Here is a Goog.. Ear.. today of the array, on the left 6KW pretty much facing E elevation about 82
degrees, operational Mar 2018. At the top 18KW sort of south maybe you can get the orientation,
May 2013, need to check elevation. In the middle 5.5KW facing sort of E, also 5.5KW facing sort
of W, elevation of 74 deg from flat Nov 2013.

There is a plan to revise everything to the newest design, hopefully before the treated wood 2
faced rots. Bruce Roe.


GearthJun20.jpg

[J.P.M.]

If you want actual equipment here, it is a lot of numbers. Or maybe just putting down a more idealized
version would work. My installer was none too careful about what was South, the west end of the S lot
line is 36 feet farther south than the east end. It was 2 years before the trash was cleaned up, I found
the stakes, and marked out the area in 100 foot squares. So only the 2018 construct is precise, but the
other facing and elevation angles can be measured out. All my E-W facing panels are mounted on a
10% grade, one end being nearly 7 feet higher than the other. I believe that is irrelevant, long as the
elevation angle is measured on a true vertical angle.

There are at least 4 different brands of panels, either 250W 60 cell or 275W 72 cell. All that could be
tracked down, along with time in service. Inverters are a pair of Fronius IG Plus 7.5KW running since
May 2013, increased loading in Nov 2013 and again in Mar 2018. Much of the current setup is slated
to be upgraded, if the foundation gets dug.

With good weather for a couple days, I will probably be working all out on the 225 foot trench, maybe
can pick up the missing numbers in unfavorable weather. If you stop by, I have extra shovels. The
trencher is working, but tough going here with all the rocks. Bruce Roe

Bruce:

To your statement: "To compare two designs, they both should be simulated the same way, such as by PVWatts. Comparing a theoretical system simulation to an actual systems performance is not apples to apples.", I'd like to compare apple to apples in terms of inputs to the greatest extent possible. Any help you could graciously provide would be appreciated.

When modeling an existing system to a given set of inputs (a process that's very similar to something called "rating" an existing boiler or other heat exchanger for different service in the power generation business I came from), using inputs that reflect reality with respect to the existing equipment as much as possible improves the probability that the model will reflect what an existing system will do.

For residential PV systems, simulated weather and irradiance conditions for the site will be the same for any modeled system.

Modeling is not an exact science and probably doesn't need to be. However, the more and better information on the existing system I have, the higher the probability will be that the model will produce numbers that reflect reality.

As for array azimuths, a google shot of the property is probably close enough. I suppose I could find one by rooting around in your zip code but you could save me the trouble.
for the rest of it, PVWatts will only need array sizes and the inverter sizes an array feeds, and array tilts. I'll also need approx. times of the year when you change tilts. The rest of the particulars I listed would be used for the SAM input.

Because you use 15 kW of inverters, I intend to use the same 15 kW in any modeling of a south facing array. I expect that will under-invert any south facing array that is modeled and sized to produce your stated annual output.

As for shading, I believe you once provided a graph that showed approx. times of shade. Any further information you might provide on shading with respect to dime of day/time of year would be helpful.

As for stopping by, without a baseball season, I don't plan on being in the Midwest anytime soon for my 1X/yr. pilgrimage to the friendly confines of Wrigley Field. But if I was, my offer was semi-serious. Having mostly Irish genes, and what's not Irish is Scot, kind of analogous to what male dogs usually do to fire hydrants, I'm sort of genetically predisposed to digging holes if there's dirt and a shovel nearby.

[bcroe]

Thank you for your response.

I wholeheartedly and very strongly agree with your 1st opinion with respect to modeling. Before my last post I was wishing I had the particulars about your arrays/system so I could do some modeling comparisons.

To that end, before I go any further into discussing the rest of your above post, if you post all the relevant input PVWatts requires including inverter AC size pertinent to your arrays, I'll run the PVWatts analysis and compare it to a run of a south facing array at 42 deg. tilt and sized so that PVWatts produces the same output with the same inverters and total inverter size for both systems. Please include the azimuths and also seasonal tilt angles and the approx. dates of tilt changes. I'll keep the single orientation array that's compared to yours at 42 deg. tilt.

I'll even go you one better. Give me module maker, model/size and approx. date when they entered service, and inverter model(s) for SAM. I'll use the same data for both systems for the SAM input.

To the extent we agree to trust models' output, we all might learn something.

If you want actual equipment here, it is a lot of numbers. Or maybe just putting down a more idealized
version would work. My installer was none too careful about what was South, the west end of the S lot
line is 36 feet farther south than the east end. It was 2 years before the trash was cleaned up, I found
the stakes, and marked out the area in 100 foot squares. So only the 2018 construct is precise, but the
other facing and elevation angles can be measured out. All my E-W facing panels are mounted on a
10% grade, one end being nearly 7 feet higher than the other. I believe that is irrelevant, long as the
elevation angle is measured on a true vertical angle.

There are at least 4 different brands of panels, either 250W 60 cell or 275W 72 cell. All that could be
tracked down, along with time in service. Inverters are a pair of Fronius IG Plus 7.5KW running since
May 2013, increased loading in Nov 2013 and again in Mar 2018. Much of the current setup is slated
to be upgraded, if the foundation gets dug.

With good weather for a couple days, I will probably be working all out on the 225 foot trench, maybe
can pick up the missing numbers in unfavorable weather. If you stop by, I have extra shovels. The
trencher is working, but tough going here with all the rocks. Bruce Roe

[J.P.M.]

Furthermore I built my Shoulder array "East-West orientation" based on information from this Fronius white paper.........

https://www.energymatters.com.au/ima...olar-paper.pdf

I have been very happy with the results.

I'm not really impressed with that paper. Seems there's nothing new in it and it also seems a bit slippery with statements. An example from the abstract: An unsubstantiated statement : " from the perspective of grid operators, east-west oriented grid systems are preferable to south oriented ones, as the energy is fed in more evenly throughout the day." That greater temporal distribution may be somewhat of an advantage, but it almost always comes at the price of decreased day long total output (The area under the south facing curve will be greater than the split array curve for the same total installed STC wattage). That statement, to me anyway, is written to help readers infer that greater temporal distribution is the only consideration, and I kind of doubt that's the case.
Moreover, that statement seems to be at some odds with the second sentence of the abstract which states: "Although south oriented systems are better, east-west oriented systems can also generate substantial savings.

To the first point of flatter temporal distribution, I'd suggest that the statement may apply more to large non residential power plants and less to residential roof top PV systems with access to NEM agreements.
And, if in reference to residential systems, if residential T.O.U. is in force with rates favoring off south orientations, for most common T.O.U. rates/times, there is usually one single orientation that maximizes annual array production or annual array revenue. They may be the same orientation but will probably be different if T.O.U. rates are applicable. In either case, it's one orientation.
With respect to the amorphous panels' orientation of the paper's tested arrays, besides being somewhat confusing in the way it's presented, the array slopes are rather low with the crystalline modules were tilted at 15 deg. If the alleged point of the paper is to show that off south orientations aren't too bad after all, a low slope east or west azimuth orientation is a good way to B.S. the point.
Reason: A vertical orientation will show the greatest variation in output as f(azimuth) of any tilt. A horizontal (zero deg.) slope will show the least (that is, none.) The less the slope, the less the azimuthal variation. Result: A 15 deg. east or west slope will show relatively little variation to a south facing slope due to its low tilt. Result: It looks like aa east or west azimuth array can be almost as good as a south facing array. The results at a higher slope will be less favorable.

There's lots more about that paper that seems less than academically honest to me , but this is a limited venue. IMO only, the paper looks like it comes from an inverter manufacturer and it's a masqueraded pitch to sell more and larger inverters.

[J.P.M.]

In addition since I'm running the system at its maximum with hours of clipping on very hot days I am very concerned with temps in the inverters and all electrical connections. I am particularly concerned about the molded case breakers since they can run very hot for extended periods. This advice from Schneider Electric provides guidance from the UL489 standard........

https://www.se.com/us/en/faqs/FA173839/

I monitor the heat at appropriate places throughout my system. I've replaced a few breakers that ran hot just as a prophylactic. A little prevention goes a long way especially if you're running on the high end. I also use silver conductive grease on all connections and terminations where appropriate and always used a torque screwdriver or wrench as appropriate. Never had a failure....knock on wood.

For heat in the inverters, if not so equipped forcing ventilation via fans can knock out ~ 1/3 of the temp. diff amb. to heat sink temps.

for connections, etc. running a heat balance may help identify way to keep things cooler. One common way to get rid of excess heat, especially in confined spaces is with increased ventilation or rearrangement of equipment locations.

[J.P.M.]

Enjoying all the activity and comments here. A couple of my comments.

My opinion is, to compare 2 designs, they both should be simulated the same way, such as by PVwatts.
Comparing a theoretical system simulation to a different actual systems performance is not apples to
apples. Real systems have variations both externally (weather) and internally (shading, wire loss, panel
mismatching or aging, etc) not accounted for in a simulation. My shading issues are not eliminated,
but improved a bit annually.

Inverters here are recommended to over panel up through 115% total, not 147%, of the AC rating. A more
subtle design feature here has been to observe that limit, at least approximately. It will not be precise till
I get the clipping meter functioning, I do not anticipate any panel throttling needed ahead of the inverter(s).

Ground mount array supports to my requirements are more expensive than the panels they hold. The
incremental cost of converting to a 2 sided mount is small, so the mount cost for a single orientation array
approaches being double per KW. Shared return wires help. Bruce Roe

Thank you for your response.

I wholeheartedly and very strongly agree with your 1st opinion with respect to modeling. Before my last post I was wishing I had the particulars about your arrays/system so I could do some modeling comparisons.

To that end, before I go any further into discussing the rest of your above post, if you post all the relevant input PVWatts requires including inverter AC size pertinent to your arrays, I'll run the PVWatts analysis and compare it to a run of a south facing array at 42 deg. tilt and sized so that PVWatts produces the same output with the same inverters and total inverter size for both systems. Please include the azimuths and also seasonal tilt angles and the approx. dates of tilt changes. I'll keep the single orientation array that's compared to yours at 42 deg. tilt.

I'll even go you one better. Give me module maker, model/size and approx. date when they entered service, and inverter model(s) for SAM. I'll use the same data for both systems for the SAM input.

To the extent we agree to trust models' output, we all might learn something.

[DanS26]

In addition since I'm running the system at its maximum with hours of clipping on very hot days I am very concerned with temps in the inverters and all electrical connections. I am particularly concerned about the molded case breakers since they can run very hot for extended periods. This advice from Schneider Electric provides guidance from the UL489 standard........

https://www.se.com/us/en/faqs/FA173839/

I monitor the heat at appropriate places throughout my system. I've replaced a few breakers that ran hot just as a prophylactic. A little prevention goes a long way especially if you're running on the high end. I also use silver conductive grease on all connections and terminations where appropriate and always used a torque screwdriver or wrench as appropriate. Never had a failure....knock on wood.

[DanS26]

Furthermore I built my Shoulder array "East-West orientation" based on information from this Fronius white paper.........

https://www.energymatters.com.au/ima...olar-paper.pdf

I have been very happy with the results.

[DanS26]

Take a look at this GSES Technical Paper.......Oversizing PV Arrays...

https://www.gses.com.au/wp-content/u...Oversizing.pdf

On page 2 it says......"Basic modelling indicates that the maximum benefit of oversizing is realised with an oversizing ratio of 150%".

I sized my system based on this paper.

[bcroe]

Enjoying all the activity and comments here. A couple of my comments.

My opinion is, to compare 2 designs, they both should be simulated the same way, such as by PVwatts.
Comparing a theoretical system simulation to a different actual systems performance is not apples to
apples. Real systems have variations both externally (weather) and internally (shading, wire loss, panel
mismatching or aging, etc) not accounted for in a simulation. My shading issues are not eliminated,
but improved a bit annually.

Inverters here are recommended to over panel up through 115% total, not 147%, of the AC rating. A more
subtle design feature here has been to observe that limit, at least approximately. It will not be precise till
I get the clipping meter functioning, I do not anticipate any panel throttling needed ahead of the inverter(s).

Ground mount array supports to my requirements are more expensive than the panels they hold. The
incremental cost of converting to a 2 sided mount is small, so the mount cost for a single orientation array
approaches being double per KW. Shared return wires help. Bruce Roe

[Ampster]

............ Capacity Factor is pretty useless without a lot more definition, particularly with respect to exactly how "maximum capacity" is defined.

Once a system is built the only metric that really matters is kWhs produced over a year.

[J.P.M.]

Whether one uses STC capacity or inverter capacity is not important as far as comparing one system to another, because they may have been designed with different goals in mind.

The use of a particular ratio or calculation is more a matter of whether one is useful for the person doing the analysis. No doubt different metrics provide insight into different elements of a system. It may even depend on the goals of the system design which may be financial, resource management, risk management or some combination of those.

I'm not sure the term "Capacity Factor" as it's defined and used (or misused) is a good metric for PV system design or system comparison. Historically, it was conceived for use in/for power generating facilities where the fuel source is more reliable and predictable. I believe it's use can a source of more confusion than help for PV system design and analysis - the case of systems with high DC/AC ratios being an example.

But if it is used, as a start I'd suggest it's important to state whether or not inverter capacity or system nameplate capacity is used in the denominator of the calculation.

As an alternate or modification to the metric "Capacity Factor" as the term might be applied to PV systems, I'd suggest changing the capacity factor by dividing it by a PV system's DC/AC ratio and using the more commonly used "nameplate" or STC system capacity in the denominator. Some confusion or slight of hand, or just plain deception might be avoided. Besides, what's the problem with being clear about stuff ?

Better yet, a more meaningful (but certainly not the only) metric - when discussing either energy or economic efficiency with respect to P systems anyway - is net annual plant output per installed STC kW.

(Bruce: If you're reading this, apologies about writing about you in the third person) This isn't an indictment of what Bruce has done and isn't meant as such at all. If I was near him, I might have volunteered my services to help dig holes or trench ditches, or wire something, etc., and considered the labor as the tuition of learning something. Being somewhat of a fellow eccentric, I get what he's done. More than a few of my neighbors see me fooling around with the stuff on my roof and scratch their heads. Besides, NOMB how someone allocates their resources, including time.

An example of why using inverter size for calculating capacity factor rather than system nameplate rating doesn't reveal much, if any information and may even be misleading:

Bruce has a 35 STC kW PV system. On 05/15/2019 he stated the system's latest output is ~ 28,500 kWh/yr., or ~ 28,500/35 ~ = 815 or so kWh/yr. per installed STC kW.

Without too many particulars about system design, a 20 STC kW ground mounted array facing south at a 42 deg. tilt in zip 61084 and powering 15 kW of inverter capacity will produce something like 28,600 kWh/yr. according to PVWatts. (That figure is after ~ 2,900 kWh/yr. of clipping caused by undersizing the (15 kW) inverters on a 22 STC kW system - see next paragraph).

Now let's bump that STC size by 10 % just in case Bruce's annual reported figure is for a very good year and leave the system output at 28,500 kWh/yr. to be on what's probably the very conservative side of things.

Using the 15 kW of inverter capacity with a DC/AC ratio of 22/15 = 1.467for each system, the capacity factor is essentially the same for each system: 28,500 kWh/yr/[((365days/yr.)*24hrs./day)*(15 kW inverter capacity) = 0.217.

However, which one do you think would cost less to acquire, install and operate in most any similar situation and climate ? A 35 STC kW ground mounted system with multiple orientations or a 22 STC kW ground mounted single orientation system ?

Put another way, which system do you think will have a higher probability of producing a lower LCOE - one that produces 28,500 kWh/yr/35 STC kW = 815 kWh/yr per STC kW, or one that produces 28,500 kWh/yr./22 STC kW = 1,295 kWh/yr. per STC kW ?

Bottom line, in vernacular terms, which system of the two will have the better chance of producing more long term bang for the buck ?

And, with respect to Capacity Factor, what does the above show about the usefulness of Capacity Factors without some considerations of how the system size is defined ?

For many folks with PV systems, like Joe/Jane 6 pack, the big reason they have PV at all is probably, if not solely, or at least primarily for the purpose of lowering their cost of providing electricity to their residence. I'd suggest such folks don't care as much about capacity factors which as shown here, may not reveal as much, or much of anything about as the cost of the system as they care about how much a system produces in the way of revenue or bill offset in $$.

Aside from the problems I have with definitions used, or assumed, or undefined when using capacity factor with respect to PV systems, it seems to me that Capacity Factor is pretty useless without a lot more definition, particularly with respect to exactly how "maximum capacity" is defined.

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

[Ampster]

..........
One measure I have used is Capacity Factor of the installed plant.
I have calculated my plant at ~20% capacity and using data you posted on PVOutput.org your plant runs at ~22% capacity. Your greater plant capacity is most likely due to DC/AC ratio and panel orientation.

Yes, I like to use those kind of metrics also. Capacity factors above 20% are not as usual because the sun only shines part of the time. Bruce has also used a ratio of annual production in kWhrs to inverter capacity in kW. That is a metric used by solar farms as one measure of cost effectiveness. You are correct that higher DC to AC ratios and orientation influence that. Whether one uses STC capacity or inverter capacity is not important as far as comparing one system to another, because they may have been designed with different goals in mind.

The use of a particular ratio or calculation is more a matter of whether one is useful for the person doing the analysis. No doubt different metrics provide insight into different elements of a system. It may even depend on the goals of the system design which may be financial, resource management, risk management or some combination of those.

[J.P.M.]

Bruce....since the nameplate capacity for our two Fronius 7.5 inverters is the same it is interesting to make comparisons.

One measure I have used is Capacity Factor of the installed plant. The actual calc is explained here.....

https://en.wikipedia.org/wiki/Capacity_factor

I have calculated my plant at ~20% capacity and using data you posted on PVOutput.org your plant runs at ~22% capacity. Your greater plant capacity is most likely due to DC/AC ratio and panel orientation.

On capacity factor, for plants, I'm not a fan of using such factors as have historically been used for nuke, hydro or fossil fuel fired power producing plants as I don't think the somewhat unique factors of fuel availability and siting are accounted for in a meaningful or useful way by the capacity factor.

That said, or written, I believe the capacity factor for PV is usually and commonly understood to be based on D.C or the nameplate STC rating of the power producing equipment, that is, in front of the inverter, with any factors including orientation or inverter size reductions and other things being reflected in the (lower) capacity factor.

In Bruce's case he's got 35 kW of panels. Any losses such as those from inverters or less than optimum orientations, etc. be they necessary, or unavoidable, or as as mandated by design or preference by the owner have the effect of lowering the capacity factor.

Using that 35 kW number as the facility (D.C.) capacity, the denominator of the equation for capacity factor for PV plants referenced in your attachment becomes :

(365 days)*(24 hours/day)*(35 kW) = 306,600 kWh.

Now, I don't recall what Bruce's annual production is, but if I use your number of Bruce's capacity factor of ~ 22%, and know you are using 15 kW of inverter capacity instead of the 35 kW of STC rating, I get a capacity factor for Bruce's system of : 0.22 *(15/35) = 0.094.

For some reference only, my long term (6 1/2 years anyway) capacity factor using averaged annual system output of 9,068 kWh/yr.:

(9,068 kWh/yr)/[(365 days)*24 hours/day)*(5.232 kW)] = (9,068 kWh/yr.)/(45,832) = 0.1979.

That includes an ~ 3.5 % output reduction for shading and an inverter that clips at 5,035 W. The 5.232 kW system size is based on 327 STC W/panel *16 panels.

I could have split the array into 3 orientations and in so doing reduced the annual output to ~, say, 8,000 kWh/yr. (making it somewhat analogous to what Bruce has), or with a 206 deg. az. and 31 deg. tilt for a single array have increased the annual output to ~ 9,200 kWh/yr. If I could have cut the tree down that's at a 220 azimuth to my array, I'd have boosted array output by ~ 3.5 %/yr. If I had done any of those things, the capacity factor would have been affected in direct proportion to the changes in output.

By the definition the capacity factor of a PV system is based on the DC (or the STC) system size, not the inverter rating. As I understand it, inverter capacity can only effect the numerator of the capacity factor equation, and only then to the extent that the inverter capacity is < the STC rating of the what is feeding that inverter.