Solar Warm Air Panel: End Of Term Report

An Eccentric Anomaly: Ed Davies's Blog

Not surprisingly my solar warm air panel isn't doing a lot now it's December so perhaps an update would be good.

Previously: Part 1, Part 1.5, Part 1.6, Initial Results, Window Box and PV Fan.

The PV-driven fan mentioned in the last of those posts failed pretty rapidly. It was exposed to both bright sunshine and torrential rain, either of which could have been the cause, but I performed a post-mortem examination which showed no signs of water ingress so I think it must have been over-voltage from the PV what done for it.

Throughout the rest of the summer and into September the collector gave a welcome trickle of warm air into the bathroom on any day with some sunshine in the morning. Even on days without enough sunshine to provide any significant heat to the bathroom it did allow ventilation to help dry out the room without appreciable cooling.

Since about the beginning of October, though, it really hasn't had any noticeable effect at all. Facing very slightly north of east is not good at these latitudes in the best of situations but with a big hill in the way (here's a view looking south east taken last winter:)

there's really not a lot of hope. By the time the sun is above the hill it's so far round to the south that it's shining on the panel at quite an oblique angle, then it disappears behind the house only an hour or so later.

The best that can be said is that the construction has stood up pretty well to the last few week's hail, snow and strong winds. The bottom edge of the PVC rattles a bit in the wind. More amusingly, it seems gusting winds cause the collector to "walk" away from the box:

Hardly a fundamental design flaw or difficult to fix if being serious about the thing.

I've also got a few little solar panels out in the garden:

There have been some windy nights of recent and my sleep has been a bit disturbed, not least by wondering where and in what condition these panels would be in the morning. Actually, they've been pretty much fine as they're weighed down reasonably heavily at the back:

Not completely obvious in that picture is that the rear legs are also tied to bits of metal banged or screwed into the ground which probably helps. The only casualty was a ring on the front bottom strut which connects to the strap to the corresponding rear leg, hence the bit of blue rope standing in for it for the moment.

Nominally the panels are 15 W each but my impression is that actually the maximum combined output is only about 35 W to 40 W. I've only used them in bright sunshine with a PWM charge controller which won't get the best voltage out of them anyway so it's hard to tell but the short-circuit current in bright sunshine seems to be just over 2 amps and maximum-power-point voltage is about 17 V.

The purpose of these panels for this winter is simply to act as solar sensors to allow me to collect some statistics on sunlight available around here. For this purpose I've now got them on a maximum power point tracking controller (a Morningstar TriStar MPPT). This is a ridiculous combination - the controller is designed to handle 30 to 60 times the amount of power - but it does allow convenient logging over Ethernet (I've written some node.js code to talk to it using Modbus over TCP/IP) and also allows me to familiarise myself with the device without worrying about oddles of amps flailing around.

I have to say there are a few minor disappointments with this device:

Still the basic MPPT function seems good and I expect I'll be using a few of these soon.

The panels came with a wiring harness which connects them in parallel for nominal 12 V battery charging. I've rewired them in series/parallel for nominal 24 V operation to make the best use of rotten conditions (hence the plastic bag "weatherproofing" tie wrapped to the back of the panel. The actual maximum power voltage (Vmp) seems to typically be around 33 to 34 V. Not that that'll make much difference just float charging an old sealed lead-acid battery.