The solar PV blog

I can’t quite decide if we have the new procedure for estimating the annual performance of a solar PV system totally sorted (as detailed in the draft of the PV installation guide – Edition 3). I am aware that a number of people are now looking at it in some detail, so I thought it would be useful to describe the thinking behind the revised procedure.

The guiding principles behind the update were to address the following:

a)    Update methodology to use regional solar data (the current procedure uses one figure for the whole of the UK)

b)    Revise pitch and orientation factors (smaller gradations)

c)    Overhaul shade assessment procedure

d)    Ensure new procedure is fully transparent and auditable

e)    Ensure new procedure us as straightforward as it can be – without completely sacrificing accuracy

Addressing the first two items was fairly straightforward, and the last two items were simply factors to bear in mind when reviewing the resulting procedure. The difficult area to address was the means to properly assess and account for any potential shade losses.

The current SAP procedure addresses shade by the use of a simple multiplier. The installer needs to assess “Overshading – % of sky blocked by obstacles” to derive the multiplication factor Zpv. Four options are available:

• None or very little (<20%) … 1.00 multiplier

• Modest (20-60%) … 0.80 multiplier

• Significant (60-80%) … 0.65 multiplier

• Heavy (>80%) … 0.50 multiplier

This is all well and good, but firstly there is no agreement on what really constitutes “overshading” or “% of sky blocked” when it applies to a PV system. This obviously makes it very difficult to audit an installer’s procedure and can lead to variations in opinion between installers when looking at the same site. Secondly, these steps cover a very big range of shading. This is particularly true for the modest shade option – which covers shading from 20-60%. Oh, by the way, I really hope no one is installing PV in sites with significant or heavy shade … but that’s another blog!

After some discussion and various iterations, the proposed shade assessment method uses a standard horizon chart, annotated to include loss zones. This is then used to record and account for any items that will obstruct the sun at different times of the day and at different months of the year.

The procedure described, can be summarised as follows: Standing as near as possible to the base of the array (e.g. through an upstairs window), use the sunpath diagram to plot all the objects (near and far) that obstruct the horizon. Once the horizon line is drawn, any square that is more than half affected should be counted (add up the figures for all exposed squares that are more than half covered by shade). The shade factor (SF) is then calculated using the following formula: SF = 1 – (sum of affected squares).

The procedure goes on to describe how to allow for arrays that don’t face due south and significantly also states that the whole process can be omitted, and an SF=1 used, where there is a clear horizon and no shade exists on the array.

I am not going to describe the procedure in more detail as it’s all in the draft of the guide … as are the sunpath diagrams which need to be seen to properly understand and assess the procedure. What I do want to finish with is to highlight the two considerations I flagged at the start of this blog:

 Ensure new procedure is fully transparent and auditable

 Ensure new procedure us as straightforward as it can be – without completely sacrificing accuracy

I am pretty sure that the answer to the first of these two is Yes (or at least will be yes once the document is finished and worked examples are presented).

Straightforward and sufficiently accurate? … probably. Having played around with other methods and considered if any of the steps presented could be omitted or simplified, I think the answer to the second item is also yes … but it’s always going to be a balance and perhaps someone can see something the team has missed?