improving pump setup on modcon p/s loops

[hlv]

Last fall I had a new modcon boiler serving two fin-tube baseboard zones. I have a couple questions regarding the current circulator pump set up and whether it could be improved.

The system is a Lochnivar Knight WH55 boiler with Grundfos pump (UPS15-58FC) on the primary loop and a BumbleBee pump (Taco HEC-2) with temperature delta T sensor on the secondary loop with two zone values. There is also a DHW priority loop with indirect tank.

The delta T on the primary loop is typically less than 5F even with the boiler-manufacturer-supplied Grundfos pump set to its slowest of 3 fixed speeds. Consequently, given my outdoor reset curve (which is low enough, maintaining two-hour long burns) the boiler doesn't condense when the outdoor temperature is below 45F or so.

I recently asked the installer about improving that, and he proposed installing a Grunfos Magna variable speed circulator that uses the boiler's 10vDC output to maintain a 30F delta T. It's probably not worth the four-figure installation quote to gain ~5% in heating fuel efficiency, but perhaps the replacement pump would also save some electricity(?) and of course I'd like to see the boiler operating at its advertised 95% efficiency in months other than May and October.

Also, I don't understand the point of having a delta T control on the secondary loop. Currently the Bumblee pump is set to maintain a 16F delta T. The observed effect is that the circulator runs at one of two speeds: initially at 13.4 gpm when there is a call for heat, then once the loop warms up, at a steady 6.8 gpm whether 1 or 2 zones are on. My impression is that the increased speed is overkill and does not affect the amount of heat delivered to the baseboards. Is that true?

Any thoughts on whether the installer's proposal is a good idea or are there cheaper alternatives? Also, can someone explain how maintaining a large delta T on the primary loop would affect average temperature and the delta t on the secondary loop? Should a different control mechanism (or none) be used on the secondary loop?

Thanks for any suggestions and insight!

Harry Voorhees

 

[Tom Sawyer]

If you don't have sufficient radiation in the house, no matter what speed the pump runs at, sooner or later the delta T is going to narrow because the baseboard can't give up heat fast enough.

 

[Sponsor]

 

[Mage182]

I concur with Tom. When I removed all of my fin-tube and replaced it with cast iron baseboard I noticed a huge difference in cycle efficiency and comfort. Plus it sounds like your unit is even correctly sized which is the problem most of us deal with.

 

[Dana]

The WH55 isn't likely to be ridiculously oversized for the heat load, but based on the symptom it's probably oversized for the amount & type of radiation. (The min-fire input to the thing is only 11,000 BTU/hr, which isn't exactly the roar of a fighter jet on it's take-off roll, eh? ;-) )

How much baseboard (and size/type) is there on each zone?

With more radiation on each zone the delta-Ts will increase some, and the outdoor reset curve can be lowered without causing short-cyling behavior, but it has to be done in a measured fashion.

 

[Tom Sawyer]

I love mod cons. No other boiler has caused so many problems in so little time. Lol

 

[Dana]

It's not the mod-con's fault that the installer neither did the math nor correctly figured out what the real solutions & problems are.

I'm neither a rocket scientist (though some of my projects have been or are in orbit) , nor a hydronic designer, but many of these misbehaving mod-con installations flunk the first-order approximation napkin-math analysis. At least this one isn't a 150K BTU/hr behemoth in a 1000' bungalow chopped up into five or six microscopic zones.

But that doesn't mean there aren't folks out there who do it correctly. Well designed well implemented systems don't show up on DIY help pages very often, but those with fatal and obvious flaws abound.

Knowing the zone-by-zone radiation size will be important for figuring out the appropriate solution here.

 

[hlv]

I'd like to clarify that there are no problems or comfort issues with the system, nor short cycling; I'm just interested in optimizing condensing efficiency, within the limitations of the existing distribution system.

The distribution system is 140' of finned tube baseboard (mostly 30 years old) divided into two zones of roughly equal size, one for each floor. Design-day heat loss calculation is about 43K BTU/hr (using Dana's degree-day method, and in the ballpark of the building energy model). I think the heat loss for each zone is comparable, maybe a little higher for the first floor due to open staircases.

(For some reason the plumber decided to increase the diameter of the supply and return pipe to the second floor zone, which is split, from 3/4" to 1". I think a consequence of this is that the second floor's calls for heat take priority over the first - as in 15 minutes vs. 2+ hours to satisfy. I've just tried closing the ball valve on that zone partway to see if that will balance things out better.)

I could probably dial down the outdoor reset curve a few degrees to improve condensing, but in my observation that will not improve the delta T on the boiler loop. Do you not think a slower, variable-speed primary pump would help?

 

[Dana]

More napkin math:

Min-fire output is about 10,500K, figuring 70' of baseboard on a zone that balances at 150 BTU/ft of baseboard, which takes an average water temp of about 110F. That's about as low as you can go with fin-tube baseboard to have anything like a linear output with average water temperature (maybe a bit into the non-linear range). Even at a delta-T of 3F that's well into the condensing zone, delivering something like 95% combustion efficiency there.

Expanding the delta-T to 30F would buy you maybe another 1% in combustion efficiency, but the non-linear output of the last 30 feet of fin tube on the zone would be all too apparent with a return water temp of 90-95. If you replaced the lower temp half of the zone radiation with an equivalent amount of panel radiator you would still maintain reasonable temperature balance, but not with fin-tube.

A single zone load 10K when the reset curve is delivering 110F AWT, is when the whole-house load is roughly 20K, or half the design condition load. Assuming an outside design temp of 10F and a heating/cooling balance point of 65F, that's a 55F delta-T when the load is 43,000 BTU/hr, so your load ramp is about 782 BTU/hr per degree-F. That means the load is at the fairly radiation balancing 20K, when the outdoor temp is 65F - (20,000/782) = 39F. Below that outdoor temps you'll need water temps to rise a bit.

At design condition you're looking at 43,000BTU/hr over 140' of fin tube, or about 300 BTU/ft. That's going to take an average water temp of 135F. With a 20F delta-T you'd still be in condensing mode, but is it really that important? At your average mid-winter outdoor temp of ~30F your heat load is about 27,400 BTU/hr which would be about 200BTU/ft, which can be delivered at an AWT of a bit under 120F, which even at a very low delta-T is still in the mid-90s for combustion efficiency. It looks like tweaking the reset curve as low and still as it can go WILL get you there, even without expanding the delta-T

On a bang per buck basis rather than spending money on fancy pumps to bring the system efficiency up, spend it on insulation. Looks like you have about 6 miles of exposed distribution plumbing down there that is one big convector/radiator dumping probably more than 1-2% of the boiler output into what appears to be an uninsulated rubble foundation. Insulating the distribution plumbing with R4 pipe insulation would be in order.
Any plumbing that is going to hit temps north of 150F (which is surely the case for the indirect loops) can't use the cheap closed cell foam pipe insulation, but fiberglass can be used anywhere. Box store 1/2" wall fiberglass pipe insulation is only about R2 and INSANELY overpriced. At 1" you'd be doing pretty well. If you can't find it at a local plumbing supply house, there are online vendors who price is more reasonably, but measure the O.D. of the pipe carefully, and buy only as much as you need.

The fieldstone foundation can be safely insulated as well. It's not clear if MassSave would subsidize much more than insulating & air sealing the band joist, but that would be extremely cost-effective on it's own merits. If you're interested in pursuing the foundation insulation stuff I can walk you through the weeds on that a bit. If history is any guide, between air infiltration and conducted heat loss the foundation losses are on the order of 10-15% of the whole house load, once you have it down to 43K @ 10F, but it's sometimes north of 20%. If you have the Manual-J to refer to, you may have some guidance on that.

 

[Tom Sawyer]

None of that makes much difference really because by the time you get the whole thing sorted out, the boiler will be on its last legs.

 

[Dana]

None of that makes much difference really because by the time you get the whole thing sorted out, the boiler will be on its last legs.

None of it?

Even if the boiler craps out next week the distribution plumbing &/or foundation insulation will be viable for several decades, and will VERY MUCH make a difference (a bigger difference in most homes than going with a 96% AFUE mod-con vs. an 85% AFUE cast-iron beast.)

Whatever is heating the place, taking a 10-15% uptick in building performance buys more comfort and efficiency than boiler type.

 

[hlv]

Thanks for the replies.

Dana, your comment suggests that increasing the delta T on the primary loop would increase it in the radiation zones, but is this necessarily the case? Can’t the flows within primary and secondary loops operate at rates to maintain a higher delta T on the primary loop than on the secondary loop (say 30F on primary, less than 20F on secondary)? I assumed that was the point of the modulating primary pump.

Thanks for the napkin math, although my figures are less optimistic than yours. For design day (0F in Lawrence), I had estimated a 325 BTU/hr/ft heat load which required 155F average emitter temp, following Siegenthaler’s advice to discount the 15% “heating effect factor” used in rating baseboards. I set the curve a bit higher in order to raise the indoor temperature above the thermostat lower setpoint, adding 15F for half delta T and pickup = 170F supply temp. Even that was lacking for the downstairs zone on design day morning so I bumped it up to 180F. Balancing the zone flows may have improved that and perhaps I can dial down the curve now. OK, maybe a few miles of pipe insulation would help too.

But even with the original outdoor reset curve, we’re talking 135F supply temp, and ~130F return (given actual primary delta T) at 35F outdoors. No condensing below that outdoor temp. Given those numbers, what do you reckon another 10 degrees drop in return water temperature would yield in efficiency? Eyeballing the condensing temperature curve, I’d guess 5% at most, depending on how much of the more of the heating season during which condensing would occur (say, down to 25F).

Let me follow-up on your cellar insulation suggestions in a separate post.

FWIW, this boiler with indirect HW is saving 20-25% over the circa 1980 Teledyne-Lars boiler and standalone gas hot water tank that it replaced. The savings exceeds the numbers predicted by the boiler efficiency ratings, as Roy Collver quantifies in his recent course on Heatspring. Sealed combustion seems to be key; the cellar is warmer now due less cold air getting sucked in.