New System Questions

[J.P.M.]

Even then, in most cases you get almost the same benefit from two arrays - one SW, one SE - in the same footprint. (And with the cost of solar running about 70 cents a watt, it's no more expensive.)

I've been hearing that for some time now, and unless some unusual circumstances intervene, I continue to question the validity of that statement for most locations.

Assuming ground mounted arrays where you get to pick orientations, haul up PVWatts and try this: Using 1 kW array size, run 3 orientations, 135 deg., 180 deg. and 225 deg. , all tilted at latitude.

Compare the outputs.
Compare: Which array of the three has the most output ?
Compare: How does the total of two, 500 Watt arrays, one facing 135 deg., the other one facing 225 deg. compare to the output of one, 1,000 Watt array ?

Any comparison I've made has the single, south facing array producing slightly more than sum of the two, 500 Watt arrays at 135 and 225 deg. orientations. There may be some locations where that may not hold, but I've yet to find one, or even conjure one up, except such as, for example, where a south facing mountain/building/obstruction shades an array between 1100- 1300 hrs. solar time.

Particulars of each application like shading, available space, etc., or, in the case of a roof location, may make one orientation better than the others, and that's usually the case, but, speaking in general terms and of production only, not revenue considerations for tariffs such as T.O.U., for most locations that have a viable solar climate, one orientation is almost always more productive than two in terms of annual output per installed kW, and that orientation is usually and mostly south facing.

For off grid applications, because of battery charging requirements and/or use patterns etc., there may be some value to split/several orientations to level out hourly production a bit or produce more output at off solar noon time(s), but that's particular to each application, rather than a general precept, and if grid tied, mostly a moot point as far as production times are concerned.

Even with most T.O.U. tariffs, one orientation is usually more profitable in terms of revenue, at least as most currently constructed T.O.U. tariffs seem to be working.

Bottom line: For most grid tie applications and locations, given the options of single or multiple orientations, choosing multiple orientations on the premise that such an arrangement will produce the same or greater output per installed kW on an annual basis than a single orientation is probably not a good assumption, is probably not correct or at least questionable for most applications and locations, and therefore perhaps not the best place to begin a design.

I also question the wisdom of assuming the cost of a split array in terms of $$'s and complexity/reliability will come in the same or less than for a single array. I'm sure the footprints will also be different.

[jflorey2]

I've been hearing that for some time now, and unless some unusual circumstances intervene, I continue to question the validity of that statement for most locations.

Assuming ground mounted arrays where you get to pick orientations, haul up PVWatts and try this: Using 1 kW array size, run 3 orientations, 135 deg., 180 deg. and 225 deg. , all tilted at latitude.

Compare the outputs.
Compare: Which array of the three has the most output ?
Compare: How does the total of two, 500 Watt arrays, one facing 135 deg., the other one facing 225 deg. compare to the output of one, 1,000 Watt array ?

I did so for my location and here's what I get:

1kW facing south tilted at my latitude - 1901 kwhr/year
1kW southwest (225 deg) - 1847
1kW southeast (135 deg) - 1749

Now, since we're talking about the same available space, both of those 1kW arrays have to be reduced to 700 watts (because that's what will fit in the same area with two tilted arrays.) Now we get 1293+1224=2517 kwhr/year for both arrays, which is 32% more than the single array.

[J.P.M.]

I did so for my location and here's what I get:

1kW facing south tilted at my latitude - 1901 kwhr/year
1kW southwest (225 deg) - 1847
1kW southeast (135 deg) - 1749

Now, since we're talking about the same available space, both of those 1kW arrays have to be reduced to 700 watts (because that's what will fit in the same area with two tilted arrays.) Now we get 1293+1224=2517 kwhr/year for both arrays, which is 32% more than the single array.

No it won't. Your 2,517 kWh/yr. will be ~ 70% of the output of two each ,1 kW arrays, one facing 135 deg. and the other facing 225 deg., which will be 3,596 kWh/yr. total.

And a 2 kW south facing array will have 3,802 kWh/yr. output, which is more than 3,596 kWh/yr., for 2 X 1 = 2 kW of STC PV.

Maybe the south facing projected area of a 135 deg. orientation will only be ~ 70.7% of that of a south facing array, but and unless I'm misinterpreting what you write, that doesn't translate to array output in any meaningful way or any way that's useful - at least not as you seem to be writing.

Unless you're referring to an approximate relative area projected into a south facing plane, I'm not sure I understand your reference. If you are referring to an array's projection into a south facing plane - that makes no sense. If you think so, see any vendor and tell them you want a 30 % discount for a 45 deg. off south array and note their response.

All other things being equal, you'll wind up paying at least as much and my guess is more for the two, 1kW arrays at 45 deg. off south orientations and get less annual output than for a 2 kW south facing array.

If I understand you correctly, and following your logic, an example: Going from the general case of projecting any array's area into a south facing plane to get it's output equivalence to a south facing array, considering an east or west array orientation (90 or 270 deg.), what would be an equivalent output using your method ?

[thundercowpoke]

The space they take up is irrelevant for me. They won't be blocking my view and I am on a large acreage. If built to its planned dimensions the array would be 50' long by 20' tall setting on 6' legs. My biggest concern is keeping the wind from damaging it. It would not be very hard to incorporate hydraulic rams (driven by my tractor) to fine tune it. Doubt that the extra kW would justify the cost though but I'm doing this for the same reason people make hot rods and custom cars, because they can.

[jflorey2]

No it won't. Your 2,517 kWh/yr. will be ~ 70% of the output of two each ,1 kW arrays, one facing 135 deg. and the other facing 225 deg., which will be 3,596 kWh/yr. total

Right.

We were talking about trackers for limited-footprint installations. As Inetdog said, they were "only remotely justifiable if you only have space for one small array of panels (which would fit on one tracker) and cannot put any more panels in place period."

So if you have limited roof area on a south facing roof then two arrays - SW and SE - placed on that footprint will give you considerably more energy than a single flat array oriented south on that same footprint - 32% more in fact. (And is still better than a tracker in the same place, since there are no moving parts.) However, if you are not limited in roof area, then laying those two arrays flat (and oriented S) will increase energy by 41%, which is better overall.

[cebury]

Wasnt the only exception to this, as JPM noted, is with a few uncommon TOU plans that greatly reward the early and (more likely) later hours. Ie. If peak pricing is far far greater than midpeak during miday and the hours are shifted just right to reward the 160/200 directions?

[J.P.M.]

Right.

We were talking about trackers for limited-footprint installations. As Inetdog said, they were "only remotely justifiable if you only have space for one small array of panels (which would fit on one tracker) and cannot put any more panels in place period."

So if you have limited roof area on a south facing roof then two arrays - SW and SE - placed on that footprint will give you considerably more energy than a single flat array oriented south on that same footprint - 32% more in fact. (And is still better than a tracker in the same place, since there are no moving parts.) However, if you are not limited in roof area, then laying those two arrays flat (and oriented S) will increase energy by 41%, which is better overall.

One more time, unless I misunderstand, what you are writing is incorrect. If you have limited roof space, a south facing orientation will be the most effective use of that space in terms of annual PV electrical energy production. Other orientations or combinations of orientations will yield less annual production, not only per m^2 of panel, but also per m^2 of available footprint. Also, because of self or adjacent row shading, single axis trackers will probably also produce less output per m^2 of footprint.

Also, and depending somewhat on latitude, and separation distance, and tilt, and even though probably somewhat minor (but perhaps a consideration that would need optimizers or micros to address), SE and SW arrays may well have additional early morning and late afternoon penalties as they shade one another a bit at certain times of the year and as f(solar azimuth angle, arry, tilt). This effect will be more pronounced in the case of space limited applications where I'd assume the south ends of each array might be almost touching.

What PVWatts is trying to tell you is that, for fixed arrays, plain and simple, SE and SW orientations having the same total area and tilt as one south facing array will, with all else being equal, produce less electricity over identical periods than the single south facing array with the same tilt.

[J.P.M.]

Wasnt the only exception to this, as JPM noted, is with a few uncommon TOU plans that greatly reward the early and (more likely) later hours. Ie. If peak pricing is far far greater than midpeak during miday and the hours are shifted just right to reward the 160/200 directions?

Not really. I was writing only about production, not revenue. The T.O.U. reference with respect to orientation was meant as parenthetical. However, with what most T.O.U. tariffs seem to look like at this time - that is, usually one peak period per 24 hrs. with no peak on weekends/holidays, it seems to me that one array orientation might usually be better than more than one, with the optimum orientation being the one that maximizes the revenue the array produces for the T.O.U. tariff used and which is also going to be f( user use patterns and times).

All of which, BTW, increases the uncertainly of future revenue from PV arrays for owners as POCO's fool with rate times for whatever reason(s), and making the idea of "optimum orientation" for an array working against a T.O.U. tariff a bit murkier.

But, that's more off topic than this whole foolishness of off south orientations being more productive and cost effective than south facing. Unless required by the application, or certain very unusual circumstances, they ain't,, and repeatedly saying they are won't change that, as the math and solar energy engineering confirm.

[jflorey2]

One more time, unless I misunderstand, what you are writing is incorrect. If you have limited roof space, a south facing orientation will be the most effective use of that space in terms of annual PV electrical energy production. Other orientations or combinations of orientations will yield less annual production, not only per m^2 of panel, but also per m^2 of available footprint. Also, because of self or adjacent row shading, single axis trackers will probably also produce less output per m^2 of footprint.

What you may be missing from my scenario is that while you can fit one square meter of panels in one square meter of available roof footprint, you can fit 1.4 square meters of panels in the same footprint if you tilt each half of the array 45 degrees away from the plane of the roof. This is ugly; you end up with a funny-looking peak in your array. It is, however, a much better option than a tracker in the same footprint (IMO.)

[J.P.M.]

What you may be missing from my scenario is that while you can fit one square meter of panels in one square meter of available roof footprint, you can fit 1.4 square meters of panels in the same footprint if you tilt each half of the array 45 degrees away from the plane of the roof. This is ugly; you end up with a funny-looking peak in your array. It is, however, a much better option than a tracker in the same footprint (IMO.)

I understand that. One large, monolithic array facing south, tilted at 45 to the horizontal located at 45 deg, latitude on a flat and horizontal roof will, in all likelihood produce more than any other single or combination of fixed arrays of any other orientation(s), and that includes a (smaller by 1*cos 45) monolithic array parallel to the horizontal roof, not only because the tilted array is physically larger, but also because the larger array has a larger projected (intercepted) to the beam portion of solar radiation when integrated over a year. That's from solar 101.

It's also pretty much why, when the goal is to maximize annual production, it usually winds up, when all the models and design are done, that the arrays be as normal to the equator as possible within some local optimization for local conditions of weather patterns, tilted toward the equator within a few degrees or so of the local latitude. That's one common optimum, and while a good goal, not always 100 % achievable.

As for tilting an array to a different orientation than the roof it sits on, a design solution done usually for some appeal to aesthetics, but in no small measure for wind design considerations, as well as greater annual production, panels on flat roofs (for example) are often and usually tilted toward the equator and arranged in separate rows (sometimes called a "sawtooth" array or pattern), with those adjacent rows given enough separation so that the rows closer to the equator do not shade the panels farther from the equator.That works well for aesthetics and makes the wind design less onerous in terms of required building modification, not to mention array structural design. But, the no free lunch law gets partially invoked by limiting the actual panel area to something like (as a 1st approx.) the area of the flat roof - not the projected area, but the flat dimensioned roof area of the tilted array ( so, a 50m^2 flat roof will only allow something like 50m^2 or so of tilted panels once necessary increases in array spacing to accomodate shading is accounted for) with the result that not much, if any, increase in panel area is achieved by practical measures taken when placing panels on nominal, or mostly horizontal structures.

My points are that for the same fixed array area, the most annual production for PV takes place - leaving out albedo effects - per m^2 of array when that array faces mostly south at approx. the latitude of the location. Other fixed orientations will yield less on an annual basis, at least for grid tied PV.

Sometimes folks confuse the steadier, day long output that intentional array variation may give (and may be necessary for the application) as having the same or more production than one, same area south facing tilted at latitude array. They (steady output and max. production) are not the same. Less "bell curve shaped" or flatter output as well as higher tilts for increases in winter production/less summer overproduction for solar thermal applications or others, may be a good choice for specific applications, but for most folks who have PV, a single south facing array is better for annual production (but again, not necessarily the best orientation for T.O.U. revenue production).

For ground mounts with few, if any area size limits, for a fixed array, and after doing all the modeling, the maximum annual output most often happens with a single orientation - and that orientation is one where the panels face mostly south and at a tilt within a few degrees of the local latitude, with large(r) arrays having several rows of panels spaced so as not to shade one another, primarily, but not entirely a winter consideration.

[reader2580]

100 KWh per day would not be unheard of with electric heat. My house previously was blowing through an average of about 136 KWh a day due to electric heat and electric hot water. House is only 1,300 square feet. House now has gas heating and hot water. My total use for 2016 was 4,537 KWh or average of 12.4 KWh per day.

[J.P.M.]

100 KWh per day would not be unheard of with electric heat. My house previously was blowing through an average of about 136 KWh a day due to electric heat and electric hot water. House is only 1,300 square feet. House now has gas heating and hot water. My total use for 2016 was 4,537 KWh or average of 12.4 KWh per day.

How's your gas use ? Without some serious conservation interventions ,what you live in sounds more like an energy sieve.

[Raj]

Hi Thundercowpoke,

I have a Radian GS4048A in my home backup system with a 195AH LiFePO4 battery bank (no BMS) installed 8 months now, and it works well. So I agree with jflorey's recommendation on the Outback Radian and the LFP batteries. However, if you are going to upgrade the Radian inverter at some point, then I would go with the GS8048A. The GS4048A is not upgradeable to a GS8048A for some reason that I couldn't get out of them other than "it's not an upgrade path".
If you do go with LFP batteries, and there are many advantages to doing so IMO, make sure to read up on them on this forum first. They need to be properly balanced and are particularly sensitive to over charging and discharging.

Good luck with your project,

Rick

[thundercowpoke]

Thank you for the input. I have been doing a lot of work on the size and configuration of what the final system will look like. I am looking at 40, 320W mono panels configured in five strings each running to an Outback FM80 going to a 800 ah 48V battery bank feeding a Radian GS8048A with a 20kW Generac natural gas generator. The system will be ground mounted but because of the size the mount will also serve as a structure to house the batteries, generator and other assorted items. We have also engineered a hydraulic tilt adjustment for the array so we can take advantage the different sun orientation over the course of a year. We still won't be able to track east to west. Because I will be doing the bulk of the work myself, with the assistance of a qualified electrician, the whole system can be built for under $50K. I don't need any permits or blessings from any government agency to do this and since I am in the middle of nowhere there is no consideration for the fire department since it would take over an hour for the nearest volunteer FD to get here.

Doing the financial spreadsheet it will be cheaper than on grid power because of the install cost of grid power. My effective electric rate for a 5 year pay out is $1.20/kWh assuming 20 kW/day production/usage for solar and $1.46/kWh for grid power at that usage. Taken out to a 20 year pay out solar is $0.42/kWh with the cost of three battery change outs and $0.61 for the grid power assuming no rate hikes(unlikely).

At that point the array would likely be worn out and in need of replacement but if I put $250/month (what I would pay in grid power) in an account for maintenance the equipment could be replaced as needed. This is where the grid power cost finally are cheaper than the solar. That being 20-25 years down the road I am betting on better technology by then and that electric rates will increase.

[SunEagle]

Thank you for the input. I have been doing a lot of work on the size and configuration of what the final system will look like. I am looking at 40, 320W mono panels configured in five strings each running to an Outback FM80 going to a 800 ah 48V battery bank feeding a Radian GS8048A with a 20kW Generac natural gas generator. The system will be ground mounted but because of the size the mount will also serve as a structure to house the batteries, generator and other assorted items. We have also engineered a hydraulic tilt adjustment for the array so we can take advantage the different sun orientation over the course of a year. We still won't be able to track east to west. Because I will be doing the bulk of the work myself, with the assistance of a qualified electrician, the whole system can be built for under $50K. I don't need any permits or blessings from any government agency to do this and since I am in the middle of nowhere there is no consideration for the fire department since it would take over an hour for the nearest volunteer FD to get here.

Doing the financial spreadsheet it will be cheaper than on grid power because of the install cost of grid power. My effective electric rate for a 5 year pay out is $1.20/kWh assuming 20 kW/day production/usage for solar and $1.46/kWh for grid power at that usage. Taken out to a 20 year pay out solar is $0.42/kWh with the cost of three battery change outs and $0.61 for the grid power assuming no rate hikes(unlikely).

At that point the array would likely be worn out and in need of replacement but if I put $250/month (what I would pay in grid power) in an account for maintenance the equipment could be replaced as needed. This is where the grid power cost finally are cheaper than the solar. That being 20-25 years down the road I am betting on better technology by then and that electric rates will increase.

I am not saying you are incorrect. I am just trying to figure out how you came up with that $1.46/kWh cost for grid power?

What would be your $/kWh rate for grid power today?