Murphymok,
You did not clarify whether or not your TD traps are discharging to atmmosphere, or to the condensate return header. If it is the former, 500 traps at even a small flow rate, notwithstanding TD traps are notorious steam wasters, letting out "X" amount of steam with each load of condensate that they let out, could mount up to a fairly large sized mass loss.
Go time a few traps, get the number of times per hour they pass flow, and look at our trap data, and see how much they are rated to pass at 200 Psi delta P, (again assuming that they are discharging to atmosphere) and multiply that number by 500, and see if you get anywhere close to your loss rate.
Now, are your steam flow measurement flow devices at the user end pressure compensated. Notwithstanding that you have minimum pressure loss (which tells me in a run that has 500 traps, that you have some pretty slow velocities) your flow measuring devices might have a compensation error in them.
Also, have you any idea, or any way of measuring what the quality of the steam is at the measurement point at the users end. You may be delivering the steam, but it may just be sloppy wet, (poor quality), which also would throw off the measurement, if not properly compensated for.
Your measurement device is probably calibrated for D&S (dry and saturated) steam.
The system that you describe to me would indicate that steam quality could be a problem.
rmw