Recommended Design Practices

IT IS EASY TO DESIGN A SYSTEM

If you want to design a customized system, the process is fairly simple. You should think of the load as being supplied by the stored energy device, usually the battery, and of the photovoltaic system as a battery charger. Initial steps in the process include:

A SIMPLE EXAMPLE

A sketch or back-of-the-envelope estimate of the system's size is often useful for purposes of discussion—either with management about how reasonable a particular application for photovoltaic would be or when talking to a system designer/supplier about the estimated cost for a system. The following example is just that: a back-of-the-envelope technique. It does not account for such design-specific details as the percent the system will be available. For example, a critical communication link might require 99.999% availability, whereas a street light might be fine at 98%. This technique will hit pretty close to the 98% mark.

The sample system is for a street light. The light is 30 watts and will be expected to run all night year-round. Low-wattage, high-efficiency lights are the types that make sense for photovoltaic systems. To see the effect of differing climates, we have designed a system for San Diego and another for Seattle.

FIRST: Design for Worst Case

For this example, the worst case is easy to determine. The load is greatest in the winter, which is the worst case for the load, and the sun shines least at this time, the worst case for the resource. In some instance, the worst case for the load is the summer and worst case for the resource is the winter, requiring you to perform two designs and then to select the one system that will carry the load through both summer and winter. For this example, we assume that the lights are needed for sixteen hours a day in winter. Therefore, the total daily energy requirement is 30 x 16=480 watt-hours a day.

SECOND: Throw in a Fudge Factor

At this point, we multiply the load by 1.5 to account for several factors that would be handled individually in a detailed design. Some of the factors accounted for by this method are all the system efficiencies, including wiring and interconnection losses as well as the efficiency of the battery charging and discharging cycle, and allowing extra capacity for the photovoltaic system to recharge the batteries after they have been drained to keep the load going in bad weather. After the multiplication, the load is figured to be 480 x 1.5=720 watt hours.

THIRD: Determine the Hours of Available Sunlight

Most solar resource data are given in terms of energy per surface area per day. No matter the original unit used, it can be converted into kWh/m2/day. Because of a few convenient factors, this can be read directly as "sun-hours a day." For example, in the publication A Comparison of Typical Meteorological Year Solar Radiation Information with the SOLMET Data Base (Albuquerque: Sandia National Laboratories, SAND87-2379), San Diego is shown to receive 4.6 kWh/m2/day in December on a fixed surface at latitude tilt (that is, tilted 32.73o up from horizontal). This information is available in other publications and is in the often-referenced typical meteorological year (TMY) database. What it means is that the referenced titled surface will receive the equivalent of an average of 4.6 full sun-hours in December. For the sake of comparison, we also look at Seattle, where the corresponding number is 1.2 sun-hours a day.

As an aside, capturing energy in winter can be enhanced by tilting the surface at a higher angle than latitude. A latitude tilt gives the best energy capture for the entire year, but circumstances may dictate that a different tilt be used. A few figures for San Diego illustrate this point. At latitude tilt, insolation for December, January, and February averages 4.94 sun-hours a day, the summer months average 6.35, and the annual average is 5.79 sun-hours a day. If the tilt is increased 15o, the winter energy increased to 5.29, summer drops to 5.73, and annual drops to 5.68.

Similarly, if the tilt is reduced by 15o, the winter energy drops to 4.31, summer is increased to 6.67, and the annual drops to 5.62.

Two points should be remembered from this discussion about tilt. First, changing the tilt of the array can enhance the energy collection for a certain season. Second, although changing the tilt does affect the annual amount of energy collected, it is not as great a change as one might believe. Note that a 15o tilt change only results in a 3% drop in the annual production of energy.

FOURTH: Determine the Size of the Array

The size of the array is determined by the daily energy requirement divided by the sun-hours per day. For San Diego, the size of the array is 720 divided by 4.6 or 156 watts. For Seattle, the equivalent is 720 divided by 1.2 or 600 watts. This is the size of the array. Of course, it must be modified by the size of the modules available. If 60-watt modules must be used, then you will wind up with 180 watts in San Diego. Remember, when converting calculated array to actual modules, always round up.

FIFTH: Determine the Size of the Battery

Most batteries will last substantially longer if they are shallow cycled, that is, discharged only by about 20% of their capacity, rather than being deep-cycled daily. Deep discharge or cycling means that a battery is discharged by as much as 80% of its capacity. A conservative design will save the deep cycling for occasional duty, and the daily discharge should be about 20% of capacity. This implies that the capacity of the battery should be about five times the daily load. To know the daily load, go back to the original load number before the fudge factor—that is, 480 watt hours. Add to this a battery fudge factor of about 50% to account for the efficiency of the battery discharge, the fact that only 80% of the battery's capacity is available, and the loss in efficiency because photovoltaic systems rarely operate at the battery design temperature.

The end result is that the battery design load is 480 times 1.5 or 720, which is coincidentally the same as the array's design load, but for different reasons. This is the daily energy drawn out of the battery, which is now multiplied by five to ensure 20% daily discharge: 720 times five or 3,600 watt hours. This is the battery capacity, which is usually given in ampere-hours so it must be divided by the voltage of the system; 3,600 divided by 12 or 300 ampere hours. Notice that the discussion of the battery's size is independent of the size of the array or the solar resource. In other words, the same battery size works both for San Diego and Seattle because both loads are the same and both arrays are sized to produce the same daily energy.

OPERATIONAL CHARACTERISTICS ARE A MAJOR BENEFIT

Whether a power system must have low environmental impact is specific to each application and location. It is clear, however, that photovoltaic systems exhibit the advantageous features of being silent and non-polluting, and of having no detectable visual or audible emissions. Photovoltaic systems are inherently stand-alone systems; they require no connection to an existing power source nor any supply of fuel. As such, they are less vulnerable to severe weather condition, poor means of access, and the like. If reduction in the consumption of fossil fuels is a primary concern in selecting a power system, photovoltaics is a good choice.

Photovoltaic systems have advantages over conventional power sources particularly where:

SUMMARY OF RECOMMENDED DESGIN PRACTICES

Many recommendations for producing a stand-alone photovoltaic system that will operate reliably for two to three decades are summarized here. These recommendations come from experienced photovoltaic system designers and installers. The best are based on common sense. Realizing that "the more specific the rule, the greater the number of expectations," we summarize some of the recommendations here.

Return to Photovoltaic home page

Acknowledgment and Disclaimer
Last Modified on 28 June, 2001