Announcement

Collapse
No announcement yet.

DIY grid-tie Solar install - West facing roof vs south facing ground mount

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • DIY grid-tie Solar install - West facing roof vs south facing ground mount

    We will be installing a residential solar system this summer. I plan on doing it owner-builder. I have experience as owner-builder, building my own house and shop 10-12 years ago and I am a school teacher with summers off. I am in the mountains of NE California and on PG&E NEM2.0 and plan on a system size around 6 kwh. We have the ability to cut any trees down.

    I have a lot of questions, but will start with the most pressing. Roof mounted west facing panels with some winter shading vs south facing ground mount panels with zero shading?

    The roof mount would go onto the shop, as my main residence is just too steep (10/12 pitch) and too tall (eves are 12-18' off the ground) for me to comfortably work up there. The ridge line of the shop roof points approx 185 degrees, meaning the west facing roof would be 275 degrees.

    The ground mount system would go up and behind the shop, facing south or slightly south west. We have rocky ground which makes digging difficult, but that has been over come before. I would run the wire overhead to the shop. It also has the added bonus of being able to sweep snow off and possibly easier to expand with proper planning.

    I understand that I will need a few more panels if I go West facing roof mount, but not sure if even adding more panels will produce the same as the south facing ground mount.

    What would you do?

    Attached Files

  • #2
    Assuming net metering connection, you did not say if you are planning a series string system
    or micro inverters. Any shade is devastating to a string, making micros or optimizers advantageous.
    A ground mount with no shade will maximize your output with the simplest cheapest, and most easily
    built string wiring system.

    I have a list of 46 reasons a ground mount is better, but the mount is almost always more expensive.
    One is ability to clear snow, and that may be made far easier if you plan it in advance. Another is you
    may simply make tilt changes twice a year, which increases output and almost eliminates snow removal
    efforts. Another reason is ability to select best orientation. Bruce Roe

    Sn4iE2.JPG

    Comment


    • #3
      Research what electrical plan PG&E will offer you. it might be that the west array will save you more kwh, by eating deeper into the "peak usage hours" of 2pm - 9pm .
      Will you have an EV you want to charge?
      Powerfab top of pole PV mount (2) | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
      || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
      || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

      solar: http://tinyurl.com/LMR-Solar
      gen: http://tinyurl.com/LMR-Lister

      Comment


      • #4
        Originally posted by bcroe View Post
        Any shade is devastating to a string, making micros or optimizers advantageous.
        The newer string inverters with an independent MPPT for each string handle shade about as well as any optimized or micro system. Once any section of a panel is shaded the bypass diode is activated and the unshaded sections of the panel AND string carry on at full power. This wasn't true of older inverters with parallel stings on 1 MPPT.

        Comment


        • #5
          If you have the capability, ground mount all the way. Some drawbacks but those are usually outweighed by the benefits.

          As for the discussion of array orientation and optimizing that orientation with T.O.U. POCO schedules, My experience is limited to my POCO (SDG & E). My semi-educated guess is that you have several billing schedules available from PG & E, all or most or them T.O.U. My short and respectful suggestion to you is,, and adding some to what Mike writes, look into it.

          My further guess is that you may find that an array orientation that's is more southerly than westerly oriented, say maybe 200 to 220 deg. at a tilt somewhere around 35 deg +/- some will produce the greatest $$ offset per generated kWh. But every situation is different. Before T.O.U. and in the days of tiered schedules, optimum orientation was a relatively simple matter. Now, with time variant rates, often overlaid with some tiering in the form of "credits per kWh" against published rates for what amounts to the lower tier use levels of the old tiered rates, things get complicated.

          A (hopefully) short story: Because I have a straight T.O.U. schedule without tiers available, I can run a model of what my array produces in a "typical" year, and because I can put that output on a spreadsheet of 8,760 rows alongside the T.O.U. rate for each of those 8,760 hours (in a year) multiply the two and come up with annual "revenue" my array will produce. I can then consider that ":revenue" as being available to be spent offsetting my electric bill irrespective of my usage pattern. I can use all electricity at off peak or on peak or at any time of my choosing while "paying" for it, partially or totally with my generation revenue.

          So, I did that for array orientations for my site from 90 to 270 deg. azimuths in 10 deg. increments and 20, 30 and 40 deg. tilts. The greatest offset revenue for my site is at an azimuth of ~ 210 deg. at about a 30 deg. tilt. ( ~ $467/year of bill offset per installed STC kW of PV at current rates and times). However, and to repeat, for my situation only (but I suspect others in other CA POCO territories due to a lot of similar rate structures and also rates), the orientation penalty in terms of $$ of bill offset is not too severe. Example: 180 azimuth, 20 deg.tilt gets $458/yr.; 240 deg. az., 40 deg. tilt gets $450/yr. However, 270 deg. 40 tilt gets $408/yr, all per installed STC kW.

          Your numbers will vary, my guess is a 270 roof mounted azimuth for you will be more hassle and in the end, less cost effective, or at least no more cost effective than a ground mounted with an azimuth closer to 200 deg. or so.

          Respectful suggestion: As Mike writes, comb your rate options and make them your friend. Pick one that your gut thinks works best for you. Then, run PVWatts for a few orientations and plug both data sets into a spreadsheet as described above. It sounds like a PITA but the rate data boils down to a big copy job - copying two different 168 hour weeks (1 for summer, 1 for winter) a bunch of times, and then making the weekday holidays into weekend days for rate purposes. Took me about 3-4 hours..

          Bottom line: Ground mount a mostly south facing array but read Bruce's threads/posts first or for design suggestions for things your mother never told you. If interested, optimize the array orientation as described above, but know that a 270 azimuth will: 1.) Produce a lot less electricity than a south facing array over the course of a year, and 2.) In spite of what you may hear, read or that may seem intuitive, be less cost effective when coupled with most T.O.U. rate schedules available from CA POCOS. But, you gotta' run the numbers for your situation and application.

          Welcome to the forum of few(er) illusions.




          Comment


          • #6
            Thanks for the replies. We will be on PG&E TOU Rate B - Which is not tiered.

            PG&E TOU Rate B -

            Summer Peak (4-9PM): $0.37522
            Summer Off Peak :$0.27216

            Winter Peak: $0.23774
            Winter Off Peak: $0.21894

            I will run some numbers through the a spreadsheet and see what I come up with. Is PVWatts fairly accurate?

            Comment


            • #7
              I plan on running Enphase micro inverters. I think, lol

              Comment


              • #8
                Originally posted by Coach v View Post
                Thanks for the replies. We will be on PG&E TOU Rate B - Which is not tiered.

                PG&E TOU Rate B -

                Summer Peak (4-9PM): $0.37522
                Summer Off Peak :$0.27216

                Winter Peak: $0.23774
                Winter Off Peak: $0.21894

                I will run some numbers through the a spreadsheet and see what I come up with. Is PVWatts fairly accurate?
                For my part, you are most welcome.

                On PVWatts as a model: I've found it to be mostly fit for purpose with a some caveats/things to keep in mind:

                1.) It's a model and not a predictor of performance. It's purpose is for preliminary system design. For its stated purpose, it seems fit for purpose as long as the user understands the terms used and the way it works. Without that knowledge, it can be like giving a loaded pistol to a 2 year old.
                2.) GIGO applies. Inputs that better reflect proposed system parameters and realities will allow more confidence in the modeled output.
                3.) Reading all the help/info screens gives a better understanding of the model's limitations as well as its possible use as a tool to think up some uses the model's designers may not have had in mind.
                4.) It does not, by itself, make any adjustments for shading. There are other tools to help estimate shading impacts, some models ("SAM" is one) and some physical tools -google "Solar Pathfinder" or others.
                5.) The irradiance and climatic data PVWatts uses is from NREL, and the same data as that used by a lot of other models. That data seems to be fairly representative of long term weather, including representative irradiance data, and are mostly fit for purpose. But, the dirty little secret is that most of the irradiance data for most locations, and a lot of the other weather data is synthetic, that is, not measured or observed but modeled/educated guesstimate. See the NREL TMY manual for gory details.
                6.) As you will read from the PVWatts help/info screens, and from what that info seems to imply, PVWatts modeled output will be different from actual system output. Even with model inputs that match the modeled system exactly, annual modeled output will probably be within ~ 10 % of any 365 day's total output. Any similar comparison for any 30 day output may be different by as much as +/- 30 % or so. obviously, daily comparisons of modeled vs. actual output are useless.
                7.) I've found there are other models that may provide more information and may be marginally better for some applications than PVWatts. "SAM" from NREL (== PVWats on steroids, or "TRNSYS" or other models for example. However, for most users who may not want to, or be able to get that involved in the minutia of residential PV design, PVWatts, when used according to its intended purpose is, IMO, fit for purpose, especially when considering the vagaries of weather and that most annual outputs will vary by 5-10 % year/year because of that weather variation.
                8.) FWIW, many PVWatts users use a 10 % system loss parameter rather than the 14 % default the program uses to give a better fit to observed performance. After ~ 5 1/2 yrs. of watching and measuring my system, its behavior and output like a hawk, I've found that to be true +/- a bit.

                A respectful suggestion: Review your rate plan very carefully, being aware of hooks (NBC's, or low use per kWh "credits, other stuff) that can screw things up. My POCO has a lot such things and they (SDG & E) seem to feel no compunction to be anything close to helpful or transparent about it. Getting rate analysis right and on a spreadsheet is my biggest PITA in all this.

                Add: Your choice on micros. NOMB, but I'd consider system simplicity a priority and avoid a lot of extra potential failure points that come with micros and design around shade as much as possible, which is easier (but maybe/probably still not a cakewalk) with a ground mount, and then either live with the (hopefully) small(er) and manageable shading loss, and the reduced cost effectiveness that comes with it, or as a last resort, consider optimizers and fewer (but still a lot IMO) fragile electronic components attached to the back of a bunch of panels. For me amyway, KISS as possible and as a design philosophy and priority has worked well for me throughout an engineering career. More components = greater probability of problems/failures/downtime/cost.
                Last edited by J.P.M.; 05-29-2019, 01:23 PM.

                Comment


                • #9
                  Originally posted by Coach v View Post
                  I plan on running Enphase micro inverters. I think, lol
                  What's driving your decision to go with micro inverters? The only real advantage is the lack of a need to mount a central inverter. There's a lot of confusion/misinformation about the effect of shade on string inverters. Bypass diodes are required in all solar panels. If one panel is shaded it does not 'take down the string' that one panel is shunted out of the string by the bypass diodes.

                  The cost difference is significant. Micro Inverters are ~3x more per watt compared to strings especially if you chose a ground mount option where rapid shutdown isn't required.

                  Comment


                  • #10
                    I did decide to go with Enphase micro inverters. The cost difference was less than $1k on a 6kw ground system. The main reason is 90% of the folks I spoke to in person, including the ones that will be helping me install, say to use Enphase micros. Only one said he would chose string inverters due to cost only.

                    Comment


                    • #11
                      Originally posted by Coach v View Post
                      I did decide to go with Enphase micro inverters. The cost difference was less than $1k on a 6kw ground system. The main reason is 90% of the folks I spoke to in person, including the ones that will be helping me install, say to use Enphase micros. Only one said he would chose string inverters due to cost only.
                      I'm really perplexed on the popularity of Enphase. What makes it worse is that when I ask people why they prefer Enphase I've never been given a reason that's accurate. Lower Line Losses? No. More Shade Tolerant? Not really. The most popular reason is that if one panel is shaded in a string of 14 panels that the other 13 panels have their output reduced with a string inverter... which is 100% false.

                      Comment


                      • #12
                        From what I heard(not what I know): warranty, ease of install, good product support, longevity of company, price becoming closer, monitoring.

                        Comment


                        • #13
                          If there is much distance between the panels and the meter, the cost of wiring becomes
                          important between micros and strings. Its about 700' from my most distant panel to the
                          meter. Seeing the wiring investment for the 2 types side by side would be significant. I
                          see it in other systems similar to mine. As is my wire covers the distance about half DC
                          string voltage and the rest AC inverter output. DC losses are just above 1%, AC losses
                          are over 3% at peak. There is a 4 gauge pair running out to my inverters, but to get
                          good control I would need about 3/0 gauge using micros.

                          The there is the issue of the inverters pushing high line voltage up toward the trip point.
                          This problem gets worse as the inverters get farther from the meter (located at the panels).
                          Caused me some grief with strings here, but not nearly as much as with micros. Bruce Roe

                          Comment


                          • #14
                            Originally posted by Coach v View Post
                            From what I heard(not what I know): warranty, ease of install, good product support, longevity of company, price becoming closer, monitoring.
                            Other manufactures like SMA and Solar Edge also offer extended warranties but IMO ~12 years is plenty. Have you seen the inverter that existed ~12 years ago? Just save the savings and if there's a failure at year 13 buy a smaller, cheaper and more efficient inverter that will probably have direct 5G or something crazy; I've never installed an Enphase system but I don't see how it could be easier than just running DC to a string inverter. I'll find out next month. Renvu was giving away M215s so we picked of a 4.6kW system for ~$3k too good to pass up.

                            Solar Edge and now even SMA if you have Tigo modules also offer module level monitoring and you don't need to buy a $500 gateway.

                            Originally posted by bcroe View Post
                            The there is the issue of the inverters pushing high line voltage up toward the trip point.
                            This problem gets worse as the inverters get farther from the meter (located at the panels).
                            Caused me some grief with strings here, but not nearly as much as with micros. Bruce Roe
                            I don't understand why micros would be better here... why not just put the inverters near the meter and run DC?
                            Last edited by nwdiver; 06-20-2019, 03:34 PM.

                            Comment


                            • #15
                              Originally posted by nwdiver View Post
                              I don't understand why micros would be better here... why not just put the inverters near the meter and run DC?
                              Right, strings avoid additional AC voltage at the inverter. Bruce Roe

                              Comment

                              Working...
                              X