Hi thanks for the replies. As it turned out my local
Screwfix didn't have any 16A type C's so I've temporarily fitted a 32A and it works fine.
You do not know it works until you test it ot it operates to disconnect power when a fault develops. All it is doing at the moment is allowing enough current to flow that the compressor starts. That is not its prime function.
Indeed. Of course, a 32A MCB cannot really be 'tested' but one can at least measure the circuit characteristics to determine whether it ought to function correctly in the event of a fault.
It's not totally clear as to whether, having failed to find a C16, the OP installed a B32 or C32, although I suspect the latter, which could represent a problem...
As a sort-of cross-thread aside ... we recently discussed the newly published 'maximum Zs' requirements (to ensure proper OPD function) and the general view was that very few circuits were so marginal that the 5% reduction would be of any relevance - but, if it is a C32, the case we are discussing could be an exception to that. Even under 'current regulations', achieving a Zs less than the maximum permitted of about 0.72Ω might be difficult (theoretically maybe even impossible with TN-S) so that, even if the circuit could just 'scrape through' on current criteria, the 5% reduction in permitted maximum to about 0.68Ω might render it non-compliant come July. Although the OP's 'final final circuit' (from the garage CU) is apparently only 0.6m long, one imagines that there is probably an appreciable run of cable, contributing to the loop impedance, from that garage CU to the origin of the installation.
I'm not sure whether the OP intends the 32A MCB as a temporary or permanent measure but, if the latter, and if it's a C32, he really ought to get the loop impedance measured (indeed, he should really do that if it's a B32, or if he reverts to a C16 - just to be sure).
Kind Regards, John