Solar Panel Performance Test
Among the smaller panels, we rate the Solarex MSX-18 a Best Buy. Uni-Solar’s ‘Triple Junction’ technology reduces losses due to shading, but when fully exposed the amorphous panels are not as efficient as monocrystalline cells.
Charging batteries is a never-ending problem aboard most boats. Long-term cruisers usually have several ways to replace consumed battery capacity—engine-driven alternators, solar panels, wind generators and occasionally water generators. Even daysailers and weekend cruisers may profit from a back-up means of battery charging.
Outboard-powered boats may sorely need one, and owners of auxiliaries may despise running the engine at anchor just to keep up with cabin lights, anchor light and stereo.
We’ve had solar panels on all of our boats for about 15 years now and view them favorably, more so now that wire leads are better sealed and 10-year warranties are common.
At the least, a solar panel can charge your battery while you’re away. At best, it can reduce engine running time to a more tolerable level.
In the March 1 and 15, 1993 issues, we discussed solar panels in some depth. We also tested 10 panels from seven manufacturers. Here’s a brief summary.
The photovoltaic effect was first observed in 1839 by Edmund Becquerel, a French physicist. The first single-crystal silicon device was made by Bell Laboratories in 1941. In 1958, the Vanguard satellite was equipped with a photovoltaic array with 14% efficiency. In more recent years, the US Coast Guard has begun using solar panels to charge the rechargeable batteries that power navigation aids. Elsewhere, solar panels power remote telephone stations, residential lighting and data transmissions from natural gas pipeline monitors.
Commercial solar cells are made from wafers cut from either rectangular polycrystalline silicon or single-crystal silicon (monocrystalline), or by depositing a thin film of silicon on a substrate (also called amorphous). Single crystal cells are the most efficient and most expensive to make, followed by polycrystalline cells and thin film technology.
Solar panels are expensive compared to gensets and wind generators. But their advantage is no moving parts, silent operation and safety.
The industry rates solar panels at STC (Standard Test Conditions), which is an indication of the sun’s intensity at noon at sea level on a cloudless day with the cells at 77°F. In reality, on such a day the panels will be much hotter, and higher temperatures reduce output voltage and amps. For this reason, panels should not be mounted flush to a surface, but elevated to allow air circulation on their backsides. Manufacturer specs often give figures for open-circuit voltage and short-circuit current, but what you really want to know is current at a specified operating voltage. Our tests measure volts and amps while charging a battery, a more practical figure, we think.
The panel’s angle to the sun also is critical. Amps increase according to the cosine of the angle of the panel as its surface comes closer to the perpendicular to the sun. In other words, at 80°-90° (ideal), amperage will be about twice that measured at 30°. Therefore, an adjustable mount is desirable, though you probably won’t be able to fully track the sun unless it lies loose on deck, which some multihull owners do while at anchor. A stern pulpit mount, however, may be able to rotate through much of the necessary arc, but only on one axis, unless you have a single, ball joint mount.
Similarly, shadows across the panel can drastically cut output. Because most single-crystal panels are series wired, a shadow reduces output by more than the percentage of surface area covered. Amorphous silicon cells, such as are used by Uni-Solar, have diodes that allow current to bypass a shadowed cell, and so are less affected. Unfortunately, amorphous panels, while cheaper to make and can be assembled into flexible panels, are also less efficient.
Tempered glass covers have the best light transmission factors, but plastic coatings from DuPont, such as Tedlar and Tefzel, begin at about 95.5% and, according to the company, degrade to just 93.7% after 15 years. Glass is less subject to scratching, discoloration and delamination, but is more easily cracked. We’d opt for glass if the panel was mounted away from possible causes of damage, but our Tefzel-covered Photocomm Lifesaver 50 panel has lasted five years without any problems.
Ten and more years ago, one of the most common causes of solar panel failure was corrosion due to moisture entering the laminate at the wire leads. As noted in our 1993 tests, we believe that most manufacturers are doing a much better job of sealing wire leads, most often with a waterproof junction box. We also see much less occurrence of delamination, earlier caused by the use of materials with much different coefficients of expansion.
Like batteries, solar panels connected in series (positive to negative) add voltage and parallel connections (positive to positive) add current. For example, for a 24-volt battery, you would use series-wired strings of two panels each, with those strings wired in parallel. For a 12-volt battery, the panels should be wired in parallel.
Blocking diodes are often recommended to prevent the battery or alternator from feeding current back through a damaged panel or at night. A diode also prevents a “current loop” from dropping voltage in other panels if one is shaded or damaged. The downside is that the diode’s resistance reduces voltage; coupled with inherent losses in the battery and wiring, you may lose up to 20% output. Unless your panel is a “self-regulating” panel, we recommend the use of a regulator (about $70), that, while not 100% efficient either, protects the battery from overcharging and disconnects the panel at night. (To be “self-regulating,” the panel’s current output at battery charging voltage should be less than about 1% of the battery’s capacity in amp-hours, e.g., less than 4 amps for a 400-amp-hour battery bank; often the term “self-regulating is used for low-voltage panels, but this doesn’t take into account battery capacity.)
Few panels are manufactured specifically for the marine industry and so most have aluminum frames, which provide rigidity but often have sharp corners. Some have a rubberized beading on the edge with rounded corners. Others, such as the Solarex Lite series, have no frames. Which you choose depends much on the mounting location, but unfortunately, the most models and best prices are those with aluminum frames.
Lastly, we learned that a 50-watt panel can put out more than two 25-watt panels, but only when fully exposed. A fully exposed 25-watt panel, however, can put out more than a shaded 50-watt panel. Bigger is better, but only if you can mount the panel where it will seldom be shaded.
The 1998 Tests
This year, we tested 10 panels from Siemens, Solarex and Uni-Solar, the most commonly represented manufacturers in the major discount catalogs—West Marine, BOAT/U.S. and Defender Industries. Additional panels, including the Best Buy Kyocera from our 1993 report, arrived too late for this article, but will be covered in an Update.
The 10 panels range from 6 to 70 watts. Eight of them use crystalline silicon technology; the remaining two use amorphous silicon. All but one—the Uni-Solar MBC-525—are rigid panels.
How We Tested
We set up our solar panels on a jig so that they could easily be pointed directly at the sun (or any angle to the sun) at any time. Then, with digital voltmeters, ammeters and a collection of variable resistors, we collected data to determine the amount of charging current for each panel into a battery with voltage between 10.5V (discharged) and 14.2V (fully charged) under optimal conditions.
We checked each panel for it’s drop-off in output with a cell shaded, open-circuit and short-circuit voltage, and output on an overcast day. The highlights appear in the graph on the preceding page, and in the table below.
What We Found
The most dramatic differences we found were in the two amorphous-silicon units—the Uni-Solar US-32 and MBC-525. Both exhibited much less degradation in performance when a shadow fell over a portion of the panel than did the crystalline-silicon models. Both had considerably lower efficiencies (watts per square foot) than did the crystalline silicon units when they weren’t shaded. With all the other panels, shading just one cell dropped the output by 66%-88%. With the two Uni-Solar panels, the loss was about proportional to the percentage of panel area shaded—less than 10%.
The Siemens panels we tested—the SP-70, SM-50, SM-46, SM-20, SM-10, and SM-6—all use monocrystalline cells, which make for high efficiency. The SP-70 has a lower efficiency than the other Siemens panels because the cells are more widely separated, resulting in more unproductive area per square foot of panel. The SM-6, SM-10, SM-20 and SM-46 are self-regulating; the SM-50 and SP-70 are not. Frames on all models are rugged and the junction boxes are well-weatherproofed.
Siemens’ SP-70 comes with blocking diodes to prevent loss of battery energy to the panel at night; the other units don’t supply them.
The two Solarex units we tested—the MSX-18 and the MSX-10L—are monocrystalline-silicon-cell panels, with the MSX-10L representing Solarex’s “Lite” line of panels. The MSX-18 is a solidly built, non-self-regulating panel with a weatherproofed junction box. No blocking diodes are provided. The MSX-18, like the Siemens SM-50, is a high-efficiency panel.
The MSX-10, too, is a high-efficiency panel, but we’re more than a little leery of the “lite” construction; it seems just too fragile for marine use. The cells are laminated between sheets of EVA with a stainless steel backing and Tedlar cover. Instructions warn you not to distort anything by tightening the screws that pass through the grommeted holes in the plate. Solarex appears to share at least some of our misgivings; the MSX-10L is only warranted for 5 years, as opposed to the MSX-18’s 10-year warranty.
The two Uni-Solar panels represent the only amorphous-silicon (or thin-film) units we tested. The MBC-525 is a flexible panel; the US-32 is a rigid one. Both panels make use of shade protection diodes (what the company calls “Triple Junction” technology), so that the increase in resistance of a shaded cell is bypassed by the diode, rather than acting as a series resistance for all the other cells. Both panels, however, suffer from the lower efficiencies inherent in the amorphous silicon technology.
The US-32, the more efficient of the two, requires 4.8 sq. ft. of panel to deliver a maximum of 2.05 amps whereas the Siemens SM-46 delivers 3.15 amps with a 3.5 sq. ft. panel.
The flexible MBC-525 is convenient if you don’t have a flat surface to mount the panel. If you do have a flat surface, though, the rigid US-32 is a much better choice. It features a multi-layer thin-film technology, which gives it about a 20% higher efficiency, and carries a full 10-year warranty as opposed to the MBC-525’s three-year warranty. It is also cheaper than the flexible models on a per-watt basis.
The US-32’s price is approximately the same, in terms of dollars-per-watt, as comparably-rated monocrystalline panels, but it takes up more room. There’s a junction box on the back of the US-32 panel for permanent wiring. The unit comes with a wired-in cable. Neither model is self-regulating.
United Solar says that the MBC series of flexible panels has been superseded by a new series (designated as US) that uses the same new three-layer technology as the rigid panels, and has comparable efficiencies.
If you can’t locate a panel in an unshaded area, the US-32 is worth considering. Otherwise, we’d opt for one of the monocrystalline units.
If you own a small boat with a small battery bank, perhaps with an outboard for auxiliary propulsion, and all you want from a solar panel is to keep the batteries from going flat when you’re not aboard, a small, 20-watt panel will do the job. Because the mono- and polycrystalline panels are more efficient when fully exposed to sunlight, these would be our first choice. Of those tested, we like the Solarex MSX-18 with its glass cover, robust aluminum frame and large junction box. The lower-efficiency Lite series, represented in this test by the Solarex MSX-10L (other sizes are available), have Tedlar covers and no frames. Unless weight or thickness is a strong consideration (which we have a hard time imagining), we’d pass over the Lite models, especially as they cost about the same (the MSX-18 sells for $210 from Defender Industries, as does the MSX-20L) and seem more prone to damage.
If shading is a major concern, the Uni-Solar amorphous (thin film) panels don’t lose as much as mono- or polycrystalline panels when partially shaded. But when fully exposed, they are less efficient.
Owners of mid-size boats that take only short summer coastal cruises may be served by the panels mentioned above. However, because it is likely that space exists to mount a larger panel, we’d recommend stepping up to a 40- to 50-watt panel. In our test group, we only looked at two, the Siemens SM-50 and SM-46.
Before making a purchase decision, you might wait until we publish the test results of the Kyocera panel, which we rated a Best Buy in 1993.
For live-aboards on larger boats who want solar panels to handle as much of their DC electrical budget as possible, buy as many of the largest panels you can comfortably mount. It would be hard to beat the Siemens SP-70, though other companies do make similarly sized panels.
Contacts- Siemens Solar Industries, 4650 Adohr Lane, Camarillo, CA 93011; 805/482-6800. Solarex, 630 Solarex Ct., Frederick, MD 21703; 301/698-4200. United Solar Systems (Uni-Solar), 1100 West Maple Rd., Troy, MI 48084; 800/843-3892.