PCBgogo

Hi experts,
I have made a few PCBs and used a relatively cheap Chinese reflow oven to solder the surface mount components.  I have done boards up to 100x100mm previously without any problem.  But one of my latest designs seems to fail whilst in the oven.  After the initial failure I asked the board house whether there was any issues with the batch that I had received, but their replay was that it was 5 months since receiving them and they suspected water ingress into the board causing the failure (The black board in the image - tested bare after a failed populated board).  I tweaked the design a little and had the PCB remade using a different board house.  The failure mode in the oven is pretty much exactly the same. (Red board in photo).
It appears that the board is getting locally too hot in the oven.  The area that fails is near the back of the oven, and part way through the heating cycle there is a bad smell and when the oven is opened I have a blistered board and oozed burnt board innards.
Given that this has failed on 2 boards from different PCB manufacturers it must be to do with either my design or the oven.  The only thing I can think about the design is that I haven't put a ground plane on either side of the board in this design.  A ground plane would at least spread the heat more evenly across the board and may stop this localised burning.  I have tried running a cycle of the oven with a thermocouple placed in through the draw and it seems to reflect the temperature expected in the profile, but it is possible that there is an exposed element close to the back of the oven.
Any ideas on this?

Thanks