Hi all... I hesitated to post this, because I didn't want to sound like I was being contrary to some of the excellent points being argued. But... <said sheepishly> nobody has pointed out that voltage rise may be a factor in some of the measurements here. Now, I'm not a solar expert by any means... my background, and current full-time career, is in wind energy and mechanical generators operate differently than PV arrays, I know. But I regularly see issues with voltage rise in lines, where the cabling is undersized. Voltage drop occurs where there is a load at the end of a line, which has resistance (as they all do, unless super cooled like was pointed out). SO, your 1800W heater load will cause a drop in voltage at the load/heater end of the line, as compared to the voltage measured at the line/source end.
But, in a generator configuration the opposite happens. People who haven't seen this happen have a hard time picturing it, so I use a water pump example. If the hose it too small, as the pump runs it can raise the pressure on the pump end of the system. The same happens with inverter or generator sources, as they push energy through a line... there will always be some voltage rise, due to line resistance. The goal is to make it manageable, so inverter and wind generator manufacturers have tolerance limits. It's even built into the UL 1741 and IEEE specs, so that systems don't shut down inappropriately. I know of one popular system where the tolerance is 2% but some are 5%. And, I've personally had to diagnose and resolve issues with systems (installed with wire too small) where the inverters shut down because they were "pumping" energy into a wire which was too small for the length of run... voltage would rise above the threshold limits and the system would shut down.
People don't talk much about voltage rise, because we're used to working with loads (that cause a drop, because we're measuring at the load end). But it happens in the opposite direction with generation systems... at least the ones I'm familiar with. Having said that, I have NO idea if PV output can work this way. But if you're working with charge controllers (on the DC side) or inverters (on the AC side), then this is a very real and measurable phenomenon.