Some Observations on Photovoltaic Cell Panels

by Oliver Seely

The text and photos on this page are in the public domain. Copying is encouraged!

The negative bottom line within the spread sheet ($-107.19) is a statement of our credit, but according to California law, there is no compensation to the customer for whatever credit might accrue during the year. That is, the electric company never writes us a check for our contribution to the power grid. We do, however, have to pay an annual TOU meter charge and Customer Charge.

There is the added complication of TOU schedules. TOU-D-2 is for big users. Those users pay a lower average kwh rate, but their flat TOU meter charge and Customer Charge are higher than those for customers on the TOU-D-1 Schedule. The first year, shown here, was on the TOU-D-2 schedule. At the end of the year shown, we changed to the TOU-D-1 schedule in the expectation of breaking even again at the end of the next fiscal year and paying lower annual fixed charges.

Here are the four rates charged on the TOU-D-1 schedule. Note the usurious Summer On-peak rate, which as it turns out, worked in our favor - see below:

Period	Cost per kwh ($ U.S.)
Winter On Peak	0.202
Winter Off Peak	0.142
Summer On Peak	0.504
Summer Off Peak	0.147

The spread sheet for 06/09/06 to 06/09/07 looked like this:

The credit accrued during this year was $134.00, as shown. The fixed charges came to $41.36, which gave us a considerable saving over the previous year. Most unfortunately, the electric company discontinued the spread sheet for their TOU customers several months ago. Instead we receive a short statement which outlines the applicable charges for that month. To those of us who took the time to learn how to read the spread sheet this decision puts us at a decided disadvantage to be able to estimate how we are doing throughout the year. As if to add insult to injury, the short statement offers the customer a labyrinthine summary of the rate schedule with charges for such things as the " Transmission Owners Tariff Charge," the "Nuclear Decommissioning Charge," the "Public Purpose Programs Charge," the "The Public Utilities Commission Reimbursement Fee", and the "California Alternate Rates for Energy Surcharge, where applicable." All of these fees are charged by the kW hour and I have been told by a representative that there is variability from one month to the next as to which are applied to a specific customer's bill so that even if the customer wanted to create a private spreadsheet, it would be impossible because the rate changes slightly from month to month owing to which of the above charges apply. On the other hand, an approximate rate can be determined by using simultaneous equations between pairs of months in which only one rate "season" was involved: winter or summer. Since on-peak and off-peak rates are different, one can then calculate each rate for that particular pair of months. It isn't exact, but it is close. It does however require a passage of six months through the year (summer through fall to winter) to be able to establish a credible estimate. Stay tuned. I have one month yet to go before I'll have a bead on just what I'm being charged for my electricity.

The large credit accrued has at least one misleading characteristic: it largely represents credit at the highest rate, that is, "Summer On Peak." If a rather small balance shift were to occur toward Net Load from Net Generation for this period, or possibly worse, if the Summer On Peak rate dropped significantly, a customer would risk receiving a large electricity bill. One implication of this subtlety is that an owner of a new solar electrical generation system which generates at somewhat below the level of use, might be shocked to receive a whopping electrical bill based on the inflated Summer On Peak rate ($0.335 per kilowatt hour for the TOU-D-2 schedule and $0.504 for the TOU-D-1 schedule). However, the matter of credit vs. charge cuts both ways. For the periods 2005-2006 and 2006-2007, our energy use off the grid was positive. That is, more energy was pumped in from the grid than was generated by the solar panels (2005-2006, +443 kw hours; 2006-2007, +312 kw hours), but most of this energy came to us during periods of the low billing rates and was offset by net energy generated during the high billing rate period (Summer On Peak). Had we been able to opt NOT to convert to Time Of Use billing, and had continued receiving a bill calculated at a flat rate, we would have had to pay for our net energy consumption. At $0.13 per kw hour (close to the going flat rate), our bills for those two years would have been $57.59 and $40.56, respectively). Customers at the greatest disadvantage are those who install solar panels to generate some small fraction of electricity used and then switch to Time Of Use billing. Partly for this reason, the California State Senate on May 24, 2007 and the California State Assembly on June 6, 2007 passed Assembly Bill 1714 (and approved by the Governor) which allows the owners of new photovoltaic systems during the year of 2007 to opt NOT to have Time of Use (TOU) metering to be installed. The summary of the bill reads, in part: "This bill would authorize the PUC (Public Utilities Commission) to delay implementation of time-variant pricing for ratepayers with a solar energy system, until the effective date of the rates established in the next general rate case of the state's 3 largest electrical corporations. If the commission delays implementation of time-variant pricing, the bill would require that ratepayers required to take service under time-variant pricing between January 1, 2007, and January 1, 2008, and that would otherwise qualify for flat rate pricing, be given the option to take service under flat rate or time-variant pricing."

For the current year in progress, a change in the Summer On-peak rate from $ 0.504 per kWh last year to around $0.35 per kWh this year has occurred for reasons which at this writing are not clear. That will put our domestic system at a decided disadvantage because of the excess of generated energy for the Summer On-peak periods. That is, the cumulative credit received for that period will be less than that granted last year.

All that having been said, any annual credit, whether reflecting rate disparities between summer on-peak generation and winter off-peak consumption or a surplus of generated energy throughout the year will end up making the electric company your favorite charity. "Drat and Blast!" you say. What is to be done? A customer not entirely sympathetic to the hollow-eyed plea for a spirit of giving from the executives of our public utilities needs to find alternative consumption strategies so as to bring the surplus or the credit down to zero. There are a number of amusing and intriguing possibilities which I leave to the creative genius and resourceful diligence of the reader to discover and to put into practice.

Solar Silliness

When we installed the panels on our house roof I had the expectation that we could allow them to sit there without a worry or care and to generate electricity during daylight hours for the next twenty-five years when the guarantee runs out. That they ought to be exposed to unshaded sunlight was obvious, but my early discovery that in order to achieve maximum output they need also to be rinsed periodically was an early lesson in the maintenance of solar panels. I have been more recently surprised that these two points are not fully appreciated by everyone, not even some "experts."

As more people install PV systems it stands to reason that some will make informed choices and others will not. It is with more than a little amusement then that one can find some rather large but ill-conceived installations carried out by people one would think should know better. Here is an ambitious private installation of approximately 35 kW on an apartment house in Santa Monica, California, consisting of both vertically and horizontally mounted panels. The vertical panels face southwest and do not receive direct sunlight until late each morning. Moreover, neither the vertical panels nor the horizontal panels at the right are tilted toward the south at the angle of latitude. The shadows cast by the 3 palm trees and the eucalyptus tree (right) for the better part of the day almost certainly will have an attenuating effect on the energy output; how much would be a function of the internal series/parallel circuitry but could be determined with a simulated equivalent unshaded system. There does seem to be a cleaning schedule in place judging from the blue crystalline appearance of the panels' surfaces, at left.

The shading of one's solar panels by a neighbor's trees can rise to a litigious level if one lives in California. The Solar Shade Control Act, signed by the governor in 1978, bans trees or shrubs from shading more than 10 percent of a neighbor's solar panels between 10 a.m. and 2 p.m. and includes shading on panels installed after the trees were planted if the trees grow to such a height to produce shade which exceeds that which is allowed by the law. A recent celebrated case invoking that law involves neighbors in a community near San Francisco. Neighbor A planted eight redwood trees, B, between 1997 and 1999. Neighbor C installed a 10 kW photovoltaic solar panel system, C, in 2001. Redwood trees, B, grew until their shade, D, exceeded that which is allowed by the Solar Shade Control Act. In December 2007, Santa Clara County Superior Court Judge Kurt Kumli ruled that six of the trees can remain and that the two generating the most shade must be removed. The case is being appealed. Stay tuned.













The California Department of Transportation building in Los Angeles (right) has a system of panels sandwiched in a casing of bullet-proof glass on the south face, but notice in the close-up that each rank of panels shadows the one below. Moreover, there is no cleaning schedule for the glass surface. If one could depend on frequent inundations blowing from the south then these panels would be periodically cleaned, but that kind of weather doesn't happen in southern California. We have lengthy periods without rain and when the storms do come they're more often in the form of vertical drizzles which will very definitely clean the uppermost rank of panels but do little good for the ones below.











The Los Angeles Convention Center has a system which was installed by the L.A. Department of Water and Power. The panels were placed around the periphery of the building well below the roof line (I would estimate 4-7 meters). The panels which are mounted on the east and west sides receive no direct sunlight for about half of each day. The ones mounted on the west side and shown in the photograph at the right are in the shade until early afternoon.


















A system consisting of 3872 300 watt panels (Schott ASE-300-DGF/50) yielding a rated power output of 1162 kilowatts was recently installed on the campus of CSU Fresno over Parking Lot V. The general contractor for this installation was Chevron Energy Solutions. The owner of the panels is MMA Renewable Ventures with which the campus has entered into a 20-year power purchase agreement at a starting rate of $0.16 per kilowatt hour and a 2% annual inflation adjustment. An examination of current rates paid by big users of electricity makes a rate of $0.16 per kilowatt hour appear to be a bit pricey. Note that there seems to be a slight tilt toward the south of 1-2 degrees, possibly with drainage in mind.



However, in the image at the right which has had its brightness reduced and contrast increased, the effect of such drainage where morning dew and occasional drizzles are the only sources of precipitation for several months running is a distinct residue which builds up over the cells at the lowest elevation of each set of panel segments. It is not clear at this writing if there is a program of routine rinsing in place.



A Case Study

A large system (557 kW) was recently installed on the campus of CSU Dominguez Hills by Sun Edison. There are 3279 panels, each rated at 170 watts, bringing the maximum rated power output to 557,430 watts or 557.43 kilowatts. The panels have been mounted nearly horizontally over Parking Lot 1. At our latitude of 34 degrees north they ought to have been tilted toward the south by 34 degrees if the objective is to maximize the generation of energy. At noon at our latitude on the summer solstice the sun is 10.5 degrees from the vertical. At noon on the winter solstice it is 57.5 degrees from the vertical. Assuming 0% loss if the panels are pointing directly at the sun, horizontal panels suffer a power loss of 1.7% and 46.3% at noon on the summer and winter solstices, respectively, for an average annual loss of 24%. On the other hand, under Time of Use (TOU) billing (discussed above), the On-peak period is from 10am to 6pm when the rate charged is higher and if the objective is to maximize one's $ credit the panels ought to be tilted appropriately in a southwesterly direction. Even though we often get brilliant sunlight in southern California from 7am to 10am, that time period still falls in the category of Off-peak.











But it gets worse than that. The lack of tilt means that there is no natural gravity runoff for rain or rinse water. If it appears to you that from the acute angle of view in the photo above the surface color is something other than the typical metallic blue of a silicon photovoltaic cell, you would be right. It appears (at this writing in the fall of 2006) that there has been no rinsing service to maintain maximum output. The surface has been allowed to collect the dustfall of greater Los Angeles since installation around four months ago during which time there has been no rainfall. It is not clear at this writing what the dark spots in the middle of several of the panels represent, but the buildup of dirt certainly doesn't bode well for the overall output of the panel array. It is also not clear at this writing who suffers the greatest disadvantage (the university or the power company) if the power output drops significantly due to lack of maintenance. Only knowledge of the specific billing arrangement worked out in the contract would reveal that information.

Do real data support the depressing conclusion expressed above? Well, yes, generally. On February 28, 2007, a cloudless day from 10am to early afternoon, the system on our rooftop peaked at 10:51 am with an average power output of 2271 watts over the 15 minute interval (7 minutes on either side) which bracketed the maximum of 2284 watts. Taking the theoretical maximum power output specification of these panels, the 2271 watt average translates to [2271/(18 x 165)]x100 = 76.5%. On that same day the power of the university system peaked at 12:15 pm, showing a power output of 319,841 watts. Carrying out an equivalent calculation one gets [318,841/(3279 x 170)] x 100 = 57.4%, a value diminished, I would offer, by the lack of tilt of the panels at the angle of our latitude. We are stymied at this point from looking more closely at these figures and trying to establish how much the diminished value is caused by the lack of tilt and how much by dustfall because the tilt of the domestic roof-top system is itself not ideal. One would need to observe the output of at least one 170 watt panel the normal vector of which is pointing directly at the sun at the time of maximum power by the array of 3279 panels to establish a credible attenuation of power owing both to tilt as well as dustfall.

Here is the one-year line chart of energy generated vs. date for the university system.





































Note the two discontinuities identified by the arrows. They represent the increased output following rinsing. That we are experiencing the driest year since records have been kept starting in the latter part of the nineteenth century, we've had many cloudless days. All maxima on the chart above are representative of energy output on cloudless days. Taking the highest adjacent maxima before and after cleaning, we have 10/27/2006 and 10/28/2006, 1644 kwh and 1930 kwh. The lower value is 85.2% of the upper value. Again on 3/15/2007 and 3/16/2007 we have 2222 kwh and 2599 kwh respectively. The low value is 85.5% of the higher value, suggesting that the event which triggers rinsing by the maintenance crew is a 15% drop from maximum expected value. The very low energy outputs and those at zero are unexplained. They are either outages of the panel system for part or all of the day or there was a failure of the data collection system. No explanation is available at this writing.

Conclusion

So as to gain maximum advantage from an installed system of photovoltaic panels, the following preliminary conclusions can be made. Most unfortunately, if you are not a resident of the State of California, only (2), (3) and possibly (1) make any sense. Read on:

1. If you are connected to a grid, install a system sufficiently large to generate as much energy as you consume during summer and winter periods, because if your rate varies by time of day and by season, and you take advantage of generating more energy than you consume during summer daylight hours, when the rate is the highest, and consume more than you generate during long winter nights when the rate is lower, there is no guarantee that such a rate schedule will remain to your advantage over the long term. In any case, you need to start thinking about a Plan B to use up the energy credit you build up throughout the year and possibly to install more panels if you find yourself suddenly having to pay for electricity.
2. Tilt your panels toward the south (in the northern hemisphere) or toward the north (in the southern hemisphere) at the angle of your latitude.
3. Regularly rinse your panels to keep them clean and to maximize their output.
4. If the panels meet all of your electrical energy needs, that is, if energy consumption is close to energy generation, then the decision to switch to "Time of Use" metering makes sense only if the Winter Off Peak rate is so much lower than the Summer On Peak rate that some Plan B for using up the accrued credit becomes financially appealing.
5. Don't opt for "Time of Use" metering if your panels produce somewhat less than your electricity requirement during the winter, but more than you use during the summer because a slight change in rate of one period vs. another can make the difference between an annual energy credit and an unwelcome electricity bill. Moreover, if the electric company eliminates the method you have used to track your credit/debit status by introducing a "new and improved" electricity statement and/or a change in rate for one or more periods without prior announcement, you'll be, in our vernacular, up a creek without a paddle.
6. If your panels produce only a small fraction of the electrical energy you use throughout the year then do NOT switch to TOU metering. Doing so would subject you to the inflated "Summer On Peak" rate which at this writing is on the order of three times the flat rate.

Send a message to Oliver about this page? Click here.