Here is a quote from the UK engineering recommendations, including the spelling mistake:
''If your solar PV or electrical storage system is to be connected to the National Grid, is to run in parralel with the grid, will shutdown during a powercut and is under 16A per phase (3.68kWp AC single phase, 11.04kWp AC three phase), then this is a most likely a Micro-Generator. Grid connections for Micro-Generators are carried out in accordance with Engineering Recommendation G98.
G98 is a simple connection procedure for Fully Type Tested systems under 16A per phase, systems are installed on an install and inform basis. If the Micro-Generator to be connected is installed in compliance with G98. The DNO can be notified of the installation under G98, up to 28 days after the commissioning date.''
It appears that the criteria of 3.68kW is the capacity of the system. A larger capacity (like 4kW) is not defined or approved as a micro-generator and would be classified as a Type A installation which covers 3.8kW to 1mW requiring prior approval via the G99 regulations, (despite the fact that it may be 'clipped' to a 3.68kW maximum export allowance) - a lengthy and cumbersome process.
I was also surprised to find that PV panels can get very hot. I was quoted a measured figure of 180 degrees (presumably Celsius) by my local registered installer who carried out a repair on some local roof panels during a sunny spell last summer. He used an infra-red non-contact thermometer and was measuring the heat of the underside of a PV panel. No wonder the panel was faulty. I would assume that all the soldered joints had melted!
Thinking about this some more a heat build-up is not that surprising. A good PV to electricity conversion is about 20%, the remainder of the sunlight power must either be reflected as light (probably less than 5% as the surface is black) or absorbed as heat, to be dispersed by radiation or conduction. On a 5kW medium sized array it has to continually disperse 4kw of energy. If an adequate airflow is not designed into the system the panels will get pretty hot. This is made worse by a flat roof installation (or a shallow slope) as the natural 'heat-rising' convection is not so effective as on a steeper slope. Also the aluminium supporting frame of the PV is mostly constructed like a shallow heat trap to the underside of the PV cells. (I wonder why at least some manufacturers might provide vent holes in the frame).
All this is not usually a problem concerning the roof structure under the PV array as the effect of sunlight directly shining on the roof without an array would be allowed for in the design. Apparently there is an appreciable increase in the risk of fire hazard, especially on tar-felted flat roofs.
Two other points now occurred to me. If the string array is faulty and not generating electricity, 95% of the sunlight is converted to heat.
PV panels are more efficient generators at lower temperatures.
Perhaps some more thought should be given to the installation and efficiency of solar panel arrays.