They also state that lowering the EWT brings only marginal operational cost savings, which do not justify the increased loopfield costs, a point we all agree here

But the opposite couldn't possibly be true?

Pick some 90F max EWT design, change it to 95F and the loop will be smaller. What's your point, that the resulting loopfield cost savings is NEVER a good tradeoff? Actually, don't answer that - I'm fine with just disagreeing.

Read the paper you were citing. At the time it was written, they state that reducing the EWT from 95F to 90F would save $6 annually but cost $310 more upfront. Over 50 years payback....a bit less if electrical costs increase, what they did.

reducing the EWT from 95F to 90F would save $6 annually but cost $310 more upfront.

Wow, your point is that 95F is the better choice in this case? I'm glad we now agree that 90F isn't always the right number.

You don't seem to understand that there is a difference between theory and practice. This is modeling, not the real life. In real life you need some safety margins, since it does not always plays out that way. Here is an example.

http://welserver.com/WEL0447/

5200 sqf, 35 miles SE of Buffalo, about 1000F higher elevations than Buffalo, manual J showed about 62KBTU/H, previous consumption data was inconclusive since the kept it cold and did not heat the whole house, was suspicious, had an energy rater confirm Manual J, proposed 4 ton system.

Customer warned me about cold temps, told him we would put a 6 ton system in for him at no extra charge, 6 ton loop (4800 ft of slinky at 8ft), installed in the fall of 2009 right before the heating season. At xmas the loopfield was already at 30F, at the end of the heating season the loop was at 22F, with outgoing water temps around 17-18F. Capacity was reduced, system used a lot of strip heat. Customer was concerned about the loop temps, but still happy about the savings versus propane. We were scratching our heads a bit, put a monitoring system on it.

While BUF Airport (30 miles NW), where the BIN data came from, did see -2F that winter a a low, the location of the house did see -26F as a peak low (low a few days ago was 12.1 for this month, it is the end of April!) BUF has a 6F 99% design temp. Loop recovered to 55F degrees during A/C season in the summer. For the last 2 years loop is running fine dropping down to 30F at the end of the heating season, see graph at WEL0447, performance is up. Unit uses 200 hours of supplement heat a year, costing $140. One could argue that it operates right on target.

Luckily, I ignored all the data and and violated the design rules and followed my instinct, and put in 6 ton unit instead of 4 ton. A few things came together:


1) Weather data was dramatically off just a few miles away.

2) Ground was not compacted around the loopfield pipe yet, causing it to underperform in the first year.

3) Could not inject heat into the loop during A/C season due to fall install

4) Customer was running the thermostat higher (72F), he enjoyed the comfort



The problem with modeling is that it does not account for some of those events. The 30/90F rule provides you with some safety margins which once in a while save your butt. Some here sit in front of a computer in a theoretical world, and think that a manual J and a modeling program tells you everything. If you don't think you need those safety margins, you simply have not been burned yet.

A successful example designed for higher than 30F is your support for always designing right to 30F? Sure, uncertainty/safety margins are an issue, but you just pointed out that sometimes 30F doesn't provide enough margin. And uncertainty varies on a case by case basis (imagine if his next door neighbor wanted a system from you now).

Hey everyone Jon's point is that 30/90 aren't always perfect to design by.
We get it.
....beating a dead horse here.

Posted By jonr on 25 Apr 2013 06:54 AM
And uncertainty varies on a case by case basis (imagine if his next door neighbor wanted a system from you now).

The point being that I am still shooting for 30F in any case, and this is where this system ended up being, but in most cases it is hard to know ahead of time when you need more or can afford less. The parameters needed to get to this might change, but the goal is still the same.
You have a theoretical approach, I have a practical one. We made our points. I agree with Joe, the horse is dead.