Indirect DWH Retrofit - Interesting Configuration Issue

[eriew]

When the house was built (1982) in New England, no furnace was included (heat pumps).

Later a furnace (1997) was added and a 40 gallon tank was added replacing one included originally. In January 2010, this furnace had to be replaced (leaking). Weil McLean WGO3 tank-less. The installer questioned why there was a 40 gallon tank when one was not needed, but he plumbed it up. I didn't worry too much about it at the time. A small Taco 4 gpm runs from the boiler coil to the indirect tank.

Next, the tank failed (leaking). I was planning to replace it as plug and play.

During a call to Amtrol (tank manufacturer) over a minor unrelated issue, I was told that this configuration is not practical. One-half inch lines and tankless coil, lower water temperature from the source, high-head losses (lines should be 3/4-in), all make this arrangement poor. The solution would be as the Amtrol schematics indicate, running a dedicated zone off the supply to the HEX internal to the Indirect, and back; ditch the tank-less heater altogether.

I agree with this approach for a new installation, but as I say, this was a retro-fit of a system that worked (sort of).

So the question is, could something like this work temporarily or less than optimum or would it be completely out of the question. Suppose there was an additional 10 foot head loss due to pipe friction? The pump runs longer.

According to Weil-McLean, the tank-less produces 3.5 gpm (from 40 to 140 deg with 200 deg boiler water). Amtrol advises this should be 7 gpm and they would like to see the water to the HEX hotter.

I have temporarily plumbed across the tank location and am using the tank-less for the time being.

To complete the project as recommended requires adding a dedicated zone with valves, new 3/4 inch pipe, etc.

I'm looking at having both configurations with the indirect in use during times when DHW only and not heating is required, perhaps on a timer in summer.

Comments are welcome!

[Dana]

The Amtrol should be plumbed to your HEATING SYTEM water, not the tankless coil, and the boiler controls should be configured to allow cool off between burns rather than maintainging temp for just the coil. The embedded coil is is an isolating heat exchanger to keep the potable & heating system water seperate, but the Amtrol has a similar, but MUCH LARGER internal coil that does the same thing, but with higher capacity. By not keeping the boiler hot just for heating water, it's standby losses fall dramatically, and the Amtrol's insulation is at least 2-3x that of the boiler, so it's losses are much lower. Whoever hooked it up to the coil was either an amateur or an idiot, and clearly didn't RTFM. He gave you the worst of all worlds- a forever-hot boiler with high standby losses, and a reduced first-hour rating that what you COULD get out of that combination of equipment.

Plumbed as per the Amtrol spec, and controlled as a "priority zone", the tank gets the full 100kbtu output of the WGO3 boiler whenever it's calling for heat, which is enough to run a shower 24/7 and then some, if you wanted to. If the boiler had 200KBTU of output, the Amtrol could take that in as well (but it doesn't put out that much, so don't sweat it.)

I think you're misunderstanding the gpm specs between the Amtrol requirement and the WGO3's coil. Amtrol wants you to pump 7gpm at as high a temp as the boiler will muster for high turbulence/good heat exchange in their heat exchanger. The 3.5gpm spec with of 100F rise on the embedded coil is the coil's limitation on how many BTUs it can deliver through the coil, when the boiler is 200F, and has nothingto do with the flow rate you're getting out of your pump. In fact, the coil's rating for 140F out with a 200F boiler temp is 175KBTU/hour, which is almost twice as fast as the burner can get the heat in(!), which means if you're drawing hot water that fast, the boiler is going to be dropping in temperature.

The head of the embedded coil will be many times that of the boiler & heating-plumbing when running the Amtrol as a heating zone. Given the modest output BTU of the boiler itself, the head of half-inch plumbing isn't likely to be a limiting issue- the BTU capacity of the heat exchanger in the Amtrol will be several times the output of the boiler. Whatever gpm the pump was able to push throught eh embedded coil, it'll likely be much higher with the comparatively low head of the boiler rather than the coil. The head of the Amtrol's heat exchanger is also pretty low compared to embedded coils.

With even 4gpm @ 100KBTU you'd be looking at a 50F delta-T on the boiler (whihc may be near it's stress-limits- look that up), with 130 F water coming back to the boiler from the Amtrol's heat exchanger you'd be delivering 180F water to the Amtrol, and get it's 170gallons per hour rating out of the system. (See: http://www.amtrol.com/pdf/MC10164.pdf ) At 6gpm you'd be treating the boiler nicer with a 33F delta-T and it would settle in at a different temperature range.

What model Taco is it? How many feet of half-inch, and how many elbows are we talking? A gazillion installatoins out there would use a Taco-007 and 3/4" pipe with an Amtrol, and deliver good performance out of a wide range of boiler outputs.

Whatever you set, up, it's good to set the Amtrol's setpoint high enough that the water returning to the boiler comes up to 140F or higher fairly quickly on a cold-firing. That protects the boiler & flue from corrosive condensation while running. On a cold start of the boiler on the Amtrol will initally be bringing the boiler up to temp and taking heat out of the Amtrol to do this. But the much higher thermal mass of the Amtrol relative to that of the boiler means it won't drop TOO drastically, but it could be as much as 5-10F. If you install an Intellicon HW+ or Beckett Heat Manager retrofit economizer on the boiler it will purge heat out of the boiler and into the at the end of a burn into the Amtrol rather than abandoning that heat in the boiler at temps 25-35F hotter than the tank. (It'll do similar smart-sensing during heating system calls during the heating season.)

It wouldn't surprise me if plumbing the Amtrol correctly and using an Intellicon you cut your annual oil consumption by 20% or more. A 100K boiler would be 3x oversized for my ~2200' (~3500', if you count basement areas) stick framed antique in Worcester, and it may be 5x oversized for a decently built 1980s house, meaing it runs closer to 70% for an as-used AFUE, compared to it's 83% AFUE label (which assumes ~1.6x oversizing to be at all accurate.)

And just GET RID of the notion of using the embedded coil- EVER! It's just leftover bad-idea from the days of cheap oil. If you have an indirect, it will (almost) ALWAYS provide higher summertime efficiency, and improve the overall wintertime efficiency. And with a heat-purging control on the boiler it'll do even better! See: http://www.nora-oilheat.org/site20/uploads/FullReportBrookhavenEfficiencyTest.pdf Pay attention to how the steel boiler (system #3) with heat purge fares even when 3x oversized. You can approach that level with retrofit controls even if 4-5x oversized.

[eriew]

Thank you for your comments. Here is an update on this project.

I am proceeding with the project using the suggestions from Amtrol and those offered in Reply No. 1. Yes, the Amtrol needs to be connected directly to the boiler, not to the embedded coil. The old tank was storage only, no coil; as stated previously, I'm not sure why it was installed in the first place when DHW could have been supplied directly from the embedded coil. That tank is gone now and I have the Amtrol, but temporarily the embedded coil is plumbed to provide DHW.

I am procuring parts and setting up the new system. All pipe to and from the coil will be 3/4 inch. The 1/2-inch will stay so that there will be two configurations chosen by valving (the embeded coil or the Indirect Water heater).

The circulator will have to be replaced (to small). I'm looking at the Taco 007. Hopefully this pump's curve (flow vs. head) and the system curve will intersect at a reasonable point to provide a decent flow with the friction head. Pump-Flow called me up the other day to offer me their Engineered Software pump package for $5,000. Sorry, not in the budget.

The run concerns me a bit. The tank is going to be 20 feet or more away from the boiler with bends to run around air ducts, air handlers, and stairwells that are in the way. I'm trying to shorten up the distance but still haven't been able to improve much on the previous plumbing route taken from the boiler to the old tank (using 1/2-inch Cu). Of course, the new route will be 3/4-inch. I really don't want to go to a larger pump than the Taco 007.

I most likely will need a mixing valve for anti-scald valve. Unfortunately, I did not purchase the pre-arranged setup from Amtrol so I will have to roll my own.

The biggest concern that I face is whether I will be able to get two 1-1/2 inch black-pipe caps off the supply and distribution manifolds going to existing zones off without breaking the manifolds necessary to allow the installation of the new dedicated zone loop to the indirect heater. The mechanic had trouble with this last year when they replace the 12-year old boiler. Of course, they just whipped out the Sawsall and started cutting away as time is money. I'm hoping to be able to get these threaded pipe caps off without cutting or breaking the black pipe. If anyone wants to share their thoughts on how this can be done effectively, I would appreciate some help. So far I've seen comments on using penetrating oil for several days on the threads before attempting to disassemble.

I am considering smart controllers now as suggested. They seem really nice but the are no cheap.

The good news is that this project is no longer in crisis mode as I have both heat and DHW and no leaks now. Amtrol and Taco have been fantastic in supporting the installation as I have called both of them several times. The main issue is that Amtrol really insists that the circulator take suction from the boiler supply for this new dedicated DHW loop which is the opposite way that all three existing loops are configured. They also want a swing check valve on the tank return which is also opposite of what I have now on the three heat zones. I can't seem to get Amtrol to specify a manufacturer for this (the existing checks are Taco 218's). Also, I see that pumps are available with built-in check valves, but I worry that this arrangement might not meet the requirement of a true swing-check valve that the indirect tank manufacturer requires.

Would there be any advantage to having a check-valve on both ends of thee loop (supply and return)?

Well, that's where I stand now on the project. Comments and suggestings are always welcome and will be appreciated.