Emhrys Barrell looks at the latest technology in solar panels, which are now producing meaningful amounts of power, at sensible prices
Which technology will work best for your boat?
Solar technology has been improving steadily over recent years, to the point where it can produce real, useful power on even the smallest boat, able to run a fridge, LED lights and instruments when you are away from shore power, or maintain the charge in your batteries when you are away from the boat. And the huge increase in solar generation across the world had brought prices down to reasonable levels. We will look now at the changes in the technology.
The output of a solar panel is measured in watts, being the product of the voltage times the current. The voltage output of all solar panels is dependent the amount of sunlight falling on them, and the load they are connected to. If it is unrestricted, the voltage of a nominal 12V panel can rise as high as 17-21V, depending on the conditions. The rated output will be delivered at a lower voltage than this maximum, but will still be 16-18V. However, this figure is higher than the safe charging voltage of a 12V lead acid battery.
Above 14.4V, open batteries will give off hydrogen and oxygen gas, while sealed gel or AGM batteries can be permanently damaged. If the output of the panel is small compared to the capacity of the battery bank, the internal resistance of the battery will keep the voltage down, but for most panels over 20W output you will need some form of charge controller.
The early controllers used pulse width modulation (PWM) to keep the battery voltage under the gassing point. This was simple and hence cheap technology, but the downside was that as the output current remained unchanged, you were extracting less than its maximum power from the panel. For example, a nominal 100W panel, with a maximum power voltage of 17.0V, will deliver a current of 100 ÷ 17.0 = 5.9A. If this is regulated down to a charging voltage of 13.0V, you will get an output of 13.0 x 5.9 = 77W: a 23% reduction from the rated figure.
This isn’t a problem if you’re just using the panel to maintain the charge in your batteries and counter off-load selfdischarge. However, if you want the panel to deliver the maximum useful power, a better controller is needed.
These are now available in the form of MPPT, maximum power point tracking units. You can find detailed explanations of how these work on the internet, but briefly they act as a DC:DC converter, taking the incoming volts and amps from the panel and converting them to volts that the battery requires, but maintaining the overall power that the panel is delivering in watts (less a small conversion loss). Hence it delivers more current to the battery than the panel is delivering, and overall delivers close to 100% of the output power of the panel. This gives them a significant efficiency gain over PWM controllers, up to 30%.
In the example above, the charging current will be 100 ÷ 13.0 = 7.7A: a 30% improvement.
The maximum power point is the voltage at which the power from the panel in watts is greatest. This varies with load and temperature, and the MPPT controller is continuously monitoring this point,
hence the name. Looking at the maths, it can be seen that the efficiency gain of MPPT over MWM is proportional to the difference in voltage of the battery and the panel. The larger this difference, the higher the difference in output of MPPT versus MWM. This is affected by several factors.
The first is battery voltage. Charging voltage in the bulk phase, up to 80% capacity, is normally 1.0V above the battery voltage. So a flat battery, with a voltage of 11.7V, will have a charging voltage of 12.7V. In our example above, this will mean the MPPT controller will deliver 100 ÷ 12.7 = 7.9A.
As the amount of charge of the battery increases, its voltage goes up – up to 12.5V in the bulk phase – and the charge voltage up to 13.5V. At this point the current will have reduced to 7.4A, so the efficiency gain over PWM will have reduced slightly.
The second factor is the temperature of the cells in the panel. As the temperature of a cell increases, the voltage that the maximum power is delivered at drops, and hence the advantage of MPPT over MWM reduces. This reduction becomes most pronounced as the cell temperature increases above 40-50°C. Cell temperature will be higher than ambient, by an amount that varies between panels mounted on a flat surface, and those standing in free air, but will be around 10°C. Therefore the PMMT benefit drop-off starts to become significant in ambient temperatures above 35-40°C, so only of limited overall effect in our temperate climates, but more significant as you move towards hotter climes.
The MPPT controller advantage also increases as the amount of sunlight falling on the panels is reduced by shading, partial cloud or low angles of sun, at the beginning and end of the day or in winter. This is a definite benefit on our boats, where partial shading from masts and rigging can be
Reduction in panel output
When you are making calculations as to the likely output you will get from a given solar panel, you have to bear in mind that the rated output, or the headline figure that it will be described and sold under, is the maximum that it will ever produce. This will only occur if the sunlight is falling directly on it, at 90°, in clear air.
For a panel laid flat on the ground or deck, this will only occur at midday, between the tropics, and at certain times of year. In our latitudes, the figure will always be reduced because the sun is never directly overhead, always being angled towards the south, by an amount that varies from summer to winter due to the tilt of the earth on its axis and its orbit round the sun. It is also reduced through the day, as the sun rises and sets, going from an angle of zero at dawn and dusk, and maximum at noon.
The amount of light that gets through is also affected by the amount of atmosphere it travels through. This is minimum when it is overhead, but as the angle dips to the horizon, the amount of air it has to travel through increases, and the available light is dispersed. You can observe this effect of the scattering of the light when the sun or moon dip to the horizon, and their apparent diameters increase.
The effect of the reduction due to the angle of incidence can be counteracted by tilting the panels towards the sun, which is why all solar farms have their panels angled to the south – but this is rarely practical on our boats. Some craft on rivers and canals have taken advantage of a further effect by mounting some of their panels vertically on the cabin sides, so they get not just the light direct from the sun, but the reflected light from the water surface – but again, this is a special situation.
The amount of light falling on the panels will also obviously be affected by cloud, with even wispy high cloud having some effect, right down to the darkest storm clouds bringing the current to near zero. The overall result of these reduction is that for realistic calculations, you should assume a maximum of
3-5 hours sunlight a day in summer, with around 80-90% of the rated output at midsummer, noon, falling away either side of this and towards winter.
Three types of solar panel are commonly available: rigid, semi-flexible, and flexible. Their design and manufacture have changed rapidly over recent years, driven by the increase in use worldwide for renewable power generation.
Rigid panels have a frame, which means they can be angled towards the sun, but are less robust and cannot be bent or walked upon. Semi-flexible panels can be bent to the curve of a deck, and can be walked on in soft shoes, but they must be supported, and cannot be repeatedly bent backwards and forwards or the contacts between the cells can break. Fully flexible panels can be curved to any shape and even rolled up, but their output is lower than the other types.
The panels are made up of separate cells of silicon, connected in series and parallel, and these can be either monocrystalline, polycrystalline or amorphous. Amorphous cells are commonly used in low-power applications, such as calculators, toys and small, low-output panels. Crystalline cells are made from molten silicon that is poured into a mould then sliced into thin cells, making them very efficient. Monocrystalline cells tend to have greater outputs, especially at higher temperatures, but the two are becoming similar in performance.
The cells have to be electrically connected by metallic strips, and on the earliest designs, these strips were printed on the front and back of the panels – positive on one side and negative on the other. However, the strips on the front face shielded the cells from the sunlight to a degree, and the latest designs now use back-contacts only, which improves the performance of a panel from 16-18% of the light falling on it to 20-25% efficiency.
The surface of the panels also has an effect on performance. Rigid panels will typically have glass protecting the cells, but this can reflect some of the sunlight falling on them, particularly at low angles of incidence. Various alternative coatings are used on semi-flexible panels, which protect the cells when they are walked on, but can also actually increase output at low angles of incidence by acting as tiny lenses and directing the sunlight onto the cells. The cells are all linked to a final connection block, with positive and negative outputs. If this block is mounted on the face of the panel, it can trip you up when walking across it. The alternative is to mount
Semi-flexible panels can be bent to the curve of a deck, and even walked on
the block on the back face, giving a smooth surface, but if the panel is to be mounted flat on the deck, this block will have to be recessed into it.
The panels will have to be mounted on the deck. Rigid framed types will need brackets, which means they can be angled towards the sun. This is useful on narrowboats, for instance, which can spend some time moored in one location. This will also allow air to flow behind the panels, increasing their output at higher ambient temperatures. However, they are not practical on sailing boats. For these you will have to use semi-flexible panels, fixed to the deck. This can either be done with screws and bolts, or they can be stuck down. Some panels come with peel-off sticky backing to allow you to do this quickly and simply. Otherwise you will have to use a thin contact glue or mastic.
As we have said, the semi-flexible type should not be repeatedly removed as this can break the delicate inter-cell contacts.
Connecting cables should be as large a diameter as possible, to reduce the voltage losses, and 6mm2 wire is generally accepted as the minimum for mediumpower installations. The actual joints between cable and connectors should also have a large contact area and be waterproof, and it is best to use the connectors provided by the panel suppliers. These will come marked positive or negative at the ends of each of the link wires, but as we moved through the system from panel to controller and then to the battery, we found that the polarity of each cable appeared to change.
To avoid confusion, work on the principle that the output from the panel is king, and mark all cable ends along from the positive output with red insulating tape, and from the negative with blue tape. This will avoid any possibility of cross-connection. While most controllers will accept reverse polarity, it is best not to put this to the test. When wiring up the system, always connect the battery to the controller first.
If possible, avoid mounting panels where permanent shadows from the mast or rigging will fall on them. This used to completely cut off the power output, but modern technology has considerably reduced this effect. However, it should still be avoided if you can. Similarly, do not mount them behind a window as this will also cut the current output considerably.
We tested a selection of panels of different types and outputs, from 2.4W up to 150W. This was not intended as a test of one make against another, but to investigate the different designs, sizes and equipment available, and the relative performances of the two types of controller. We also included panels that would charge your phone, tablet or laptop, this being the most basic power requirement most of us would have.
We tested the output of each panel into two identical batteries – one 10% full, with a no-load voltage of 11.7V, and one 75% full, with a no-load voltage of 12.5V.
We tested the output at midday, at the end of June, the time of maximum output of the sun in the northern hemisphere, and then at 4pm to gauge the reduction in power. We also tried a day when the sun was partly obscured by high cloud. We laid the panels flat, on a sheet of plywood, to simulate the typical installation on a deck, so the output we recorded was always going to be slightly down on the maximum possible if we had angled the panels at 90° to the sun, as even in midsummer the sun is never exactly overhead at our latitude.
We connected each through a PWM controller, and then an MPPT controller. We measured output from the panel to each controller, in volts and amps, then from the controller to the battery. The MPPT controller had a digital display of input and output parameters, but we double-checked this with our own ammeter and voltmeter.
We measured the overall dimensions of each panel, then the size and area of the cell array to allow us to calculate output per square metre.
The face of a back-and-front-contact panel, showing the connecting strips printed on the surface, and the connection block
Our test rig, with solar panel, controller, battery and clamp ammeter