A few things I'm not clear about in the 1MW "reactor 600" Fig 17 : http://pdfaiw.uspto.gov/.aiw?docid=20140326711&PageNum=14&IDKey=72E161583AAE&HomeUrl=http://appft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2%2526Sect2=HITOFF%2526u=%25252Fnetahtml%25252FPTO%25252Fsearch-adv.html%2526r=1%2526p=1%2526f=G%2526l=50%2526d=PG01%2526S1=20140326711.PGNR.%2526OS=DN/20140326711%2526RS=DN/20140326711
It doesn't show the actual position of the reactor element[s] inside the "reactor shelter" [606], although other elements like the internal reservoir [612] pumps [610] and flow meters [614] are shown. There are no pumps shown from the external tanks [620a,620b] into the reactor shelter : the text says that pumps [610] do this, though they're shown as pumping water from the internal reservoir back to the tank. (There are a total of 80 pumps! Hard to squeeze them all into the diagram.) The readings MW1a and MW1b must be from the Flow Meters into the reactor -- Tank a : 1050 to 1750 kg Tank b : 2900 to 3900 kg. I presume that the water cycle to and from the top (a) and bottom (b) tanks are run at the same time, but it's not clear why they use different amounts of water. It might be that the "shelter" either contains spare heating elements on one side, or that one side isn't fully populated. None of this matters to the COP calculation, of course. I haven't reconciled why my "conservative" COP is greater than theirs. I tried splitting the calculation in two sections, allocating the power in proportion to the separate flows. This gives COP = 12.84 and 12.83 using the reduced mass and ignoring super-heating. I get a slightly higher value for heating the water to boiling point -- 120 vs 117 kWh (they use a specific heat of 1.14 kWh/gK, mine uses a lonlinear official water table), but a smaller value for evaporation -- 143.8 kWh vs their 144.01
