Out of all the candidates in the sieve, 671 of them had a factor somewhere between 51P-52P. They were almost certainly all different factors rather than the same one. The important thing is that those candidates are proven composite (not prime) and will be removed from further consideration. Any factor returned is tested by the validator to make sure it really is a factor and then it's tested again by me later before the now-known-composite candidate is removed from the sieve.
Here are the first few factors in that range:
51000761915701909 is a factor of 163187*2^12490935+1
51001288191895927 is a factor of 200749*2^40718586+1
51005036548895321 is a factor of 238411*2^46965396+1
51006857875746229 is a factor of 99739*2^49878210+1
51010368588514211 is a factor of 163187*2^29439495+1
In the notation we use: p is a factor of k*2^n+1
Those candidates on the right hand side are now proven composite, so they don't need to be tested by LLR. We search p values sequentially and right now the candidate list is for n values of 10M (10,000,000) to 50M (50,000,000). You can see that all the candidates show an n value in that range.