Yoder, I think we may be on different pages. My boiler loop here hits the domestic HX first, then to the A/W in the furnace, then to the 200k unit heater in the shop, then the 10x20" HX before going back to the boiler, all in series. If there is P/S at the boiler, the heat loss will be the same through the building as it is now, so the temps in the underground loop would be identical regardless of how it's piped at the boiler, right? The radiant is all run in a closed system (you know how I like atmospheric radiant) off of that 10x20 HX, with all 3 zones having their own pump and mixing valve. I'll start with all 3 zones at 100 degrees as a baseline and see how it performs, that way the flow rate through the boiler loop remains the same regardless of load. P/S inside the mechanical room may have a little benefit from a no-load flow standpoint, but P/S at the boiler would only be for boiler protection if the return temps get to be too low due to the huge delta T. With all 3 radiant zones pulling at once, say 100 degrees, while the boiler loop is hypothetically still 170 into the 10x20 HX at 7 GPM, we're still looking at only 6-8 GPM through the plate on the radiant side instead of the full 15 because much of the return water will be recycled. With that, let's call it 80 degree return temps from the floor. Doesn't take a ton of new heat to inject from the plate to make it back to 100 right? So even if I'm getting 130 off the plate, should still be great plenty to feed all 3 zones while keeping the boiler loop above condensing temps. This is why I push closed systems, to keep max flow in the boiler loop. I spent many a night awake designing this system to work with what he's got, and I'm fairly certain it'll work just fine. The unit heater in the shop will almost never run at the same time as the 21 loops of radiant in there, it's only backup if he brings a semi in and wants to heat it up quick- at which point the air temp would satisfy the radiant stat and shut it down, so that's a huge weight lifted right there