I admit it: I’m a little bit behind. Quite a few trawler owners long ago added solar panels to their boats, adding many benefits—including reducing the time they spent with a noisy generator running at anchor.

Last summer, I finally decided to get it done. My objective was to reduce, if not eliminate, one of our two typical daily generator runs at anchor. We usually run the genset once in the morning and then again in the evening before bed. If I could eliminate the evening run, I’d call my solar project a success.

Of course, as an engineer, I also had some ideas about how I wanted to implement our solar system. Some of my ideas were pretty simple and obvious. Others were a bit more challenging.

I had a list of goals for the project, chief among them to get as much solar capacity as I could within the confines of the hardtop and pilothouse roof. I wanted to maintain convenient access to the other equipment that’s up there, and I wanted to avoid drilling a bunch of new holes in the fiberglass hardtop and pilothouse roof. At the same time, I wanted to maintain the look of our Selene 60, Koinonia.

And, with budget always a consideration, I intended to do most, if not all, of the installation work myself. That way, I’d save costs and be confident that the job would be done the way I wanted it.

Choosing the Right Panels

The first and most basic question was, What type of solar panels would best meet my requirements? Should I use the semi-flexible panels that mount directly to the fiberglass (or canvas), or use rigid panels that are typically mounted a couple of inches off the fiberglass?

The semi-flexible solar panels have a few compelling advantages: You can pretty much fill all of the available space with these panels, since you can walk on them to get access to other equipment; they are essentially invisible from anywhere except above the boat, since they are thin and mount directly to the fiberglass deck; and mounting them is a straightforward process. They can be screwed down, or a variety of adhesive solutions are available.

However, these panels also have some significant drawbacks. First and foremost is the problem of heat.

Solar panels lose efficiency as they get hot. The amount of energy the panels produce for a given amount of light (called irradiance) drops as the panels heat up. All the major manufacturers specify a temperature coefficient that describes how much power they lose for every degree of temperature rise. Semi-flexible panels, which are typically screwed or adhered to the fiberglass deck, have limited cooling potential in direct sunlight. By contrast, rigid solar panels can be installed with an air gap underneath. Rigid panels definitely had the advantage in this regard.

The second major disadvantage of flexible panels is cost. For the same amount of power, high-quality flexible panels would have been more than three times more expensive than rigid panels.

I chose to go with rigid panels mounted a few inches off my hardtop and pilothouse roof.

The Problem of Shading

Another important consideration is the panels’ performance with partial shading. Unfortunately, it’s just about impossible to mount solar panels on the roof of our boat without some shading from antennas, masts and other gear. And, of course, the shadows move around and change shape as the boat swings on an anchor.

Shading wouldn’t be such a big deal if the panels only lost power proportional to the amount of area under the shade. But with most panels, there’s a fairly dramatic drop in the panel’s output when even a small portion of the panel is shaded. The problem is even worse if you connect two or more panels in series, since shading of one panel affects the output of the entire string.

One panel manufacturer has done some innovative things to improve performance in partial shading. This manufacturer also published a white paper showing the results of real-world shading tests, comparing various brands of similar panels. The reported improved performance with partial shading was definitely a big plus in my book.

Panel Size and Layout

A final, but important consideration, is the “form factor” (the length and width of the panels).

Early in my research, I got drone photos looking directly down at my hardtop, and I took measurements of the space available, including obstructions such as the main radar antenna. I wanted to leave a walkway between the panels, to facilitate access to other equipment on the hardtop and pilothouse roof.

In the end, I chose 400-watt panels from Solaria. They fit my available space well, and offered considerably better performance in partial shading. I had room for four panels on the flybridge hardtop and two more on the pilothouse roof, while still providing a reasonable walkway.

Mounting the Panels

I investigated a variety of mounting systems and settled on a scheme that takes advantage of the factory-installed handrails on either side of my hardtop. My mounting system is based on running four 1-inch stainless tubes across the hardtop from handrail to handrail. Each pair of panels spans two of the athwartships mounting tubes.

To support the tubes in the center, I sourced stainless-steel standoff posts designed for hand railings. They are just the right height to support the center of the athwartships tube so it follows the curvature of the hardtop. The athwartships tubes are attached to the existing handrails with off-the-shelf Bimini-top hardware that clamps over the handrails, so I didn’t have to remove the handrails.

The pilothouse roof did not have handrails on either side, as the hardtop did, so I installed them. I over-drilled the mounting holes, filled the over-drilled holes with epoxy, and then drilled the mounting screw holes in the epoxy to ensure that water could not get into the pilothouse roof structure, even if the sealant on the screws failed. I ran a pair of similar 1-inch stainless tubes across the pilothouse roof between the new handrails with the same center standoff in the middle to support the rails.

The only custom fabrication my scheme required was welding some small tabs onto a dozen 4-foot-long aluminum angle bars—one bar for each side of the six panels. The aluminum angle bars are drilled and bolted to the panels using the factory-drilled mounting holes in the panel frames.

The tabs in the aluminum fit into standard Bimini-top slide forks, to attach the panels and brackets to the 1-inch aluminum tubes running athwartships. I had the aluminum angle pieces black-anodized, both to match the aluminum frames on the solar panels and to protect against corrosion.

A nice side benefit of this mounting scheme is that the screws can be removed from the inner pair of tabs, and the whole panel can be tilted up on the outer tabs for easy access to service and cleaning underneath. I even got some 4-foot lengths of 1-inch aluminum bar that I drilled at each end to make supports that will hold the panels securely in the raised position, to make service and cleaning easier.

Electrical Connections

With the mounting system complete, the next stage was electrical hookup.

One of the key considerations at this step is whether the panels are connected in series or parallel, or a combination of the two. Most solar charge controllers require at least a few volts of “headroom” between the output voltage from the panels (the input to the charge controller) and the highest battery charging voltage during bulk phase.

I have a 24-volt system with AGM batteries, and the bulk charging voltage is about 28.8 volts. For maximum efficiency, I needed an output voltage from the panels of at least 32 volts. If the panels produce too low a voltage, it may be necessary to connect pairs of panels in series to get a high enough input voltage to the charge controller. However, panels connected in series are more susceptible to performance loss in partial shading, so I wanted to avoid series connections if possible.

In my case, because the 400-watt panels have a maximum power voltage of around 40 volts, I had plenty of margin above the battery charging voltage, so I was able to put all the panels in parallel.

Of course, with all six panels in parallel, I had to route 12 fairly large (10 AWG) wires down from the panels through weathertight cable glands and into the boat. I routed the eight wires from the four panels on the hardtop to a pair of oval-shaped cable glands on either side of my radar mast. From there, the cables run down through the mast into a hollow part at the after end of the hardtop. There was already good access to this area, since the cabling for a variety of other electronics was similarly routed through the mast into the hardtop.

I combined the four grounds just inside the hardtop into a single, heavy (6 AWG) ground wire, and ran the four positive leads separately to a breaker/combiner box that I mounted in a locker on the flybridge. The four leads from the pilothouse panels are routed through a similar cable gland, and come to the same combiner/breaker box.

The breaker box has a separate breaker/disconnect for each panel, and the combined output from all of the panels goes to a single Victron MPPT SmartSolar Charge Controller mounted under the pilothouse settee. There was convenient access in that space to a heavy positive cable from the battery that supplies the main breaker panel.

I installed a 100-amp, ANL-type fuse between the output of the MPPT controller and the battery. The charge controller could theoretically produce as much as 90 amps of battery charging current if all six panels operated at their full capacity. While it’s unlikely that such a scenario will ever occur, a 100-amp fuse seemed appropriate. It also protects the wiring from any faults.

Early Results

I finished my solar project just as the summer cruising season was winding down, and the early results are encouraging.

In our first couple of days at anchor, I found that even on a day of broken clouds and sun, the solar array could more than keep up with the typical operating loads. We have a lot of refrigeration aboard, which is a major contributor to our load, so I wasn’t sure how realistic it would be to eliminate our evening generator run.

The answer seems—at least so far—to be that we can reasonably expect to run the generator just once in the morning to charge batteries from overnight. My hope is that on mostly sunny days (which are sometimes few and far between in the Inside Passage to Alaska), I can even stop the morning genset run early as the batteries come out of bulk charging phase, and then allow the solar panels to complete the charge throughout the day.

Overall, I’m very pleased. And now I’m wondering, Why did I wait so long?

This article was originally published in the April 2022 issue.