I'm a homeowner in MA (Bedford weather station) looking to convert from oil to gas. HW is indirect. I have 62mbh of baseboard radiation (at 180*), micro-zoned as 26/23/12, but boiler-based measurements indicate an actual design-day heat-loss of under 30mbh. I don't trust that last number enough to risk a boiler which cannot generate the full 62mbh. The current boiler is a System2000 firing 105mbh input, which has somehow withstood 28 years of short-cycling and is still flue-testing at 86%. I'm hoping for a gas system with even higher efficiency, in the manner BNL measures it (http://www.nora-oilheat.org/site20/uploads/FullReportBrookhavenEfficiencyTest.pdf). Also, I'm tired of seeing heating season reflected in my electric bill. I think that means ECM pumps, and my bank account says only as many as necessary. Searching these forums has been a real education, though probably not the easiest way to get it. I think I can probe most of the right issues and understand most of the answers. Thanks a lot. I think I'll need it: The bids I've been getting for a gas system were all at or above my current firing rate, and never addressed trying to avoid short-cycling. I'm looking to do a collaborative installation where I roughly spec the system I want and find professionals who can execute using all the good practices that I'll never know. All I need to do is figure out what that optimal system is. I'm hoping that the system can run most of the time in the range where baseboard temps, HW temps, and condensing come together. If my heat-loss data is right, I should be able to hold a comfortable design-day temperature with 140* water. And I think we would be comfortable down to 110* HW. It would be really neat if you bear out my thinking, because I would get a much better system at negligible increase in price over the usual bids (boiler, indirect, many 007 pumps). On the other hand, I do tend to overreach …. I'm currently thinking modcon and buffer to get the efficiency I want. I like the Solo110 and the KNB081 for their self-cleaning and low-head HX. My gas company on the other hand is discounting the Alpine. I would lean toward the Knight because it can handle my maximum radiation but has a 16mbh turndown. The Solo would give me faster HW, but with only one full bath I don't think I need that. I'm hoping to combine the buffer and HW by using a reverse-indirect. That would save me space, money, and standby-loss. I'm hardly being innovative here - this configuration has been in earlier forums (http://www.greenbuildingtalk.com/Default.aspx?tabid=53&aff=14&aft=74997&afv=topic among other). I like the 4-port Ergomax (http://www.ergomax.com/New-Tanks.htm), but some have questioned its insulation and build quality. I'm not sure how to pipe a Turbomax . I think the 24-33 gallon sizes would give me enough buffering, but I could afford larger if you say that it is worth it. I've lost several electric HW tanks to galvanic corrosion at the copper-steel junction. Do these units avoid that problem? What is the current thinking about whether the SH secondary should pull from the top or bottom of the tank? I know I need an ECM pump to reduce power use, and probably a smart pump to adjust for the varying active zones. But I have no idea how that translates into a choice of product. The Solo manual seems to indicate it can run with a single pump if a delP valve is used. Is that accurate and is it a good way to go? Does that also apply to the Knight, where the manual does not suggest it? Finally, I'm a little insecure about the controls. I would like to go with digital setting and reading of HW temp. If I go with the Knight, can their HW sensor attach to a "foreign" tank? It seems to me that I can ignore the HW pump output since the boiler pump will come on and make HW for all calls, and that the Knight needs no additional controls. I've learned a lot, but I'm still a newbie to hydronics and modcoms. Are my understanding and assumptions correct? Would it work? Are there better options? Are there some critical factors that I haven't mentioned which should really be specified? This is a niche case of all-baseboard and modest HW seeking efficiency. Please help.

The only way you will improve boiler efficiency is by keeping on a modcon boilerthe condensing temperature as low as possible.95 F would be great but your baseboards are not sized for that type of temperature return. If your boiler return is above 130F then no matter what boiler you select will probably be no more efficient then what you presently have installed. Nothing also has been mentioned about outdoor reset.

Thanks for the response. Sorry my post was so unreadable -- all my careful paragraphs merged when I pasted them in. My next house won't have baseboard, but this one does -- so 95* won't provide much heat. I am hoping to be able to do 120* in / 110* return in milder weather, and 140* in / 120* return for much of winter. Is that a pipe dream? I believe ODR come with all the boilers I'm considering, and I expect to be tweaking the curve.

I thought the System 2000 did heat purges on the boiler into the indirect reducing/recovering the ineffciencies of shorter cycles. What was/is your min-burn time? How many zones? Size of smallest zone?

If you have a bunch of smaller zones a single ECM pump and zone valves would likely handle the radiation end, and a 1-speed 007 (or maybe a 3-speed if you want to tweak the flow on the boiler loop) would do it.

Outdoor reset might provide slightly better comfort but would have little bearing on the efficiency of the system when the boiler is slaved to the reverse-indirect. If you set the reverse indirect's aquastat to 135-140F an the radiation flow to a ~ 20F delta-T it's just fine to set the boiler output to whatever temp it takes to provide sufficient BTUs for the domestic HW load at whatever the fixed flow is on the boiler loop and it'll be condensing most of the time and it'll have the necessary radiation temp for design condition heat.

It's the return temp, not the output temp that determines the combustion efficiency of the boiler- cranking 150-160F out is fine, so long as in operation it doesn't routinely exceed the max delta-T on the boiler. (Most can handle deltas of 50F or more, but not 100F). The temp at the bottom of the tank is what determines the return-water temp to the boiler, and during active burns it won't be much more than 5F below the temp at the top of the tank (it will be somewhat more stratified at the beginning of burns, especially during hot water draws.) The tank temp setting is then a primary determinant of you operational efficiency. Setting the tank at the lowest temp that still meets the DHW and space heating requirements is key. If it won't work for you at temps under 140F you might as well be using a cheaper non-condensing low mass water-tube boiler.

If you set the tank temp as low as 120F you may need something bigger than an ErgoMax E23 to fill a tub. With 130F water entering the boiler at high-fire a mod-con might be getting only 87%, but at lowest fire it'll be nudging 90%. With entering water temps of 120F it'll be in the 90s independent of firing rate:

http://www.mnshi.umn.edu/images/boilergraph-1.jpg

Setting the tank to 130F and the boiler output to a fixed 150F and adjusting boiler loop flow such that the firing rate is at or near min-mod when the tank's aquastat is satisfied results in very good modulation and high efficiency under all circumstances. Under heavy DHW draws by the time the tank temp at the top has dropped to 110F (about the min you'd want during a tub fill), the firing rate will be more than double what it is at the end of a burn. But setting the tank at 140F and the boiler output to 160F it'll usually be out of condensing range knocking several percent off the as-used AFUE. It'll still modulate fine, and it'll give you more peak burner output on DHW draws but the efficiency probably won't break 90%. It takes a bit of tweaking of tank & boiler temps and flows to max out the efficiency, but you can usually get pretty good performance out of it if you have sufficient radiation/baseboard to work with.

If you know or can calculate the design condition heat load and the required water temp ahead of time it helps to figure out what's the best solution. If you actually need 160F water on design day it may be more efficient to use outdoor reset and run the hot water tank as a separate zone, but since your estimated/calculated heat load is less than half the 180F rating of the fin-tube, it's very likely that you'll very likely be able to run sub-140F water year round.  If you can run 135F it can be pretty much a "set and forget" situation.

FWIW: I have an Ergomax in my system, and while the insulation seems less than specified, the rest of the design is rock-solid.  The way in which the boiler water injector is designed to create turbulence at the HW coils for better heat exchange seems right, and there's less exterior plumbing to do than with the Everhot EAs or TurboMaxes.  (Improving the insulation any of the with exterior tank-wrap or rigid foam is cheap & easy.)

The hot water output of any of them needs both a  heat trap (to keep it from convecting heat away during standby) as well as a thermostatic mixing valve (IIRC, required in MA.)

Dana, thanks for responding: I've noticed some meaty responses on your part as I've dug through the forums recently.

Yes, System2000 post-purges, at least with my 1982 controller, into the last zone to call. I bought it for its top-rated seasonal AFUE (still is according to BNL) before I understood about short-cycling. With my micro-zones the post-purge sometimes results in a noticeable room temperature overshoot. I don't know if the controller has a minimum burn-time, other than that implicit in typically having to heat up a cold HX. I fire the boiler at 105mbh input (.75 nozzle) and have a 30 gallon indirect (an unpowed electric tank) and 3 zones @ 26, 23, & 12mbh. Two boiler-based experiments indicate my design-day heat-loss is under 30mbh, so I'm hoping to be able to keep the baseboard under 140*.

Since I'm heating with oil it was irksome to see the heating season on my electric bill, so I'm shooting to avoid that in this conversion. Yes, I'm figuring on one ECM (smart ECM?) and zone valves.

Thanks for citing some specific temperatures. I'm still thinking through these, and also the appropriate control settings if I go with the Knight boiler. I need to study the Knight manual some more, and still don't have my head around how the system will respond to the various calls. But I do have a few preliminary questions:

• When the baseboard tables say I can get by at 140* on a design-day, can I translate that into operating at 150* in and 130* out, and then balance the rooms by adjusting the concentration of dust-bunnies?
• If I were to, for example, set the tank to 130F and the boiler output to 150F, would I expect the baseboard to see close to the 150F as input?
• During an infinite shower (which we don't actually take), the boiler would have to fire at or near full just to keep up. Will a 150F boiler set-point accomplish this? And if I just wash the dishes, I really only need a low firing to recover, but the boiler doesn't know this. Will HW draws cool the tank return enough to generally get good condensing, or does your 5F rule apply? I think I would actually like to modulate based on the rate of tank temperature drop, but Knight doesn't offer that option.

Could you expand on some questions about the reverse-indirect?

• There was some discussion (BM 2002 and forums) about whether drawing from the bottom of the tank might provide better mixing and be a net gain. Do you have an updated opinion on this?
• Part of the build-quality issue involved leaking at the aquastat. Are you aware of an ongoing quality issues?
• Looking at the Turbomax, where I can find extensive performance data (http://www.idcassociates.com/Produc...Tables.pdf), it seems that within the range in which I am operating (2.5gph and 60-80mbh) the HW capacity depends only on the boiler capacity but the size determines the buffering. Is that valid?

I now know to put HTML around my text. Sorry about the first post.

I you set the boiler output to 150F and the tank to 130F you'll get somewhat higher than 130F output to the radiation during an active burn, but nowhere near 150F.  It's a function of the comparative flow rates of the radiation loop vs. the boiler loop. With a tee on the Turbomax or Everhot EA you get less mixing, higher radiation water temps than with the ErgoMax, but you'll likely get somewhat better DHW performance out of the Ergomax due to the way the boiler injector is designed to induce turbulence on the coils, and is providing 100% boiler output water directly on those coils. The difference probably isn't huge but is probably measurable (I have no independent data.)

A pretty good rough estimate of boiler output is 1000BTU/hr for every 2gpm of flow, times the delta-T on the boiler.  For the infinite shower example, let's say you have 3gpm on the boiler loop (1.5KBTU/hr for every degree F between in & out), and 150F output, and you like scalding-hot 110F showers.  At the point the tank temp has dropped to your scalding-shower temp that's

(150F-110F=) 40F delta-T

x 1.5KBTU/hr = 60KBTU/hr

...by the time the temp drops  to your 110F min.

For a 2gpm shower with incoming water at 40F, and 110F water out, that's a 70F delta-T @ 1KBTU/hr per degree F, which 70KBTU/hr, a 10KBTU/hr shortfall.

But the buffering of the tank sources a substantial amount of the heat before it gets to that stage:

A 23 gallon TurboMax or Ergomax contains ~190lbs of water.  Starting out at 130F then dropping to 110F (a 20F delta-T) it delivers 20Fx190lb= 3800 BTUs to the output (above & beyond what the boiler is delivering).  That's good for only about 3.8 minutes-worth of 2gpm, 70KBTU/hr showering if the boiler never fires up. 

Assuming a 5F hysteresis on the aquastat the boiler kicks when the mid-tank temp drops to 125F (which is before the top of the tank does, in a DHW-only situation) or about 1 minute into the shower. At that point the boiler's output with a 3gpm flow and now more than 25F delta-T (the bottom of the tank will be initially colder than 125F in this scenario), you get at least 25F x 1.5K= 37.5KBTU/hr into the system from the boiler (enough to handle your space heating load with good margin, but not 2x overkill.) 

At that point the boiler is providing more than half the 70KBTU/hr drawn down by the shower and, the rate a which the tank's temp is dropping slows.  By the time the tank temp hits 115F, the boiler output is up to 52.5K, and the rate at which the tank's temp is dropping is even slower.

With a low-flow 1.5gpm showerhead it would then just keep up forever at 115F. (70F x 1.5/2 = 52.5KBTU/hr.) 

Another way to improve showering efficiency & extend showering time is with a drainwater heat exchanger.  At 2gpm and 40F incoming water a decent DWHX delivers better than 20KBTU/hr to the system (providing it's feeding both the indirect and the cold side of the shower), and it'll should pretty much keep up even with the combined load of all the baseboard zones running, the shower.  As the temperature of the water drops, the output of the baseboard also drops, and by the time the tank temp is at 110F the heat emitted by the baseboard is about half to 2/3 what it is at 130F, so the DWHX is basically covering that draw and then some.

If you're tweaking/limiting your boiler loop flow to 2.5 gallons, multiply the output numbers in the above discussion by 2.5/3= 0.833  (50KBTU when the tank is at 110F, 31KBTU/hr when the boilerstarts up with a mid-tank temp of 125F.)

Regarding where to draw from the tank and other issues: 

The ErgoMax boiler water return port is near the bottom of the tank, and the boiler output water is injected in from the top.  The heating radiation return is also routed to the bottom via a dip tube, and the heating output is at the top. This ensures that the cooler water in the tank is entering the boiler for the highest condensing efficiency, and that the hotter water is headed out to the radiation. More mixing isn't a good thing, so long as there's sufficient turbulence to get a good heat exchange on the DHW coils.  For TurboMax or Everhot EA tanks, put the tee for the boiler output & radiation draw at the top of the tank, and the tee for the boiler/radiation returns at the bottom to max out the delta-T on both the boiler and the radiation.

I have never heard of leakage at the aquastat port of any of these manufacturers.  If your water supply has hard water, just as with a tankless you'd want to plumb in some descaling ports to be able to give the potable side of the heat exchanger a rinse.  The sheet metal housing fit & finish of the Ergomax has a cheap feel to it, but functionally it looks like a slightly better thought out design from a DHW production and installation ease point of view. Inside the sheet metal it's a round steel tank with an insulation wrap, and a small amount of insulation under the top cover.  Mine had no insulation on the bottom, so I mounted it resting on a 3" thick pad of rigid foam.

And yes, the temperature & volume of the tank determine the amount of BTUs buffered,  but it's ultimately the boiler output that determines the first-hour gallons.

Correction in the simple arithmetic: The E23 has 26 gallons, not 23, so it's has ~217lbs of water in it, not 190, so between 110F & 130F it's buffering 4340BTUs, good for 4.3 minutes of 60KBTU/hr shower without invoking the boiler.

With 150F boiler output and 2.5gpm of flow, at the end of the burn with the aquastat set to 130F the firing rate at the end of burn is ~25KBTU/hr, less than half the max-output of any of the boilers in question.

During shorter draws such as dishwashing (or even high-efficiency clothes washing) the tank's temp doesn't drop enough to trip aquastat to call for heat. The hysteresis of the aquastat is one of the determining factors of the minimum burn time. IIRC those used by Ergomax have ~ 7F nominal hysterisis. In the 2.5gmp boiler loop, 217lbs of water scenario that means the min burn is 7F x 217lbs= 1519BTU, simple recovery time, no heat being drawn. At an average output of about 30KBTU/hr that takes about 3 minutes- not a disaster for a low-mass mod-con, but you'd want something longer for a high-mass boiler for efficiency's sake. Given that a low-mass tankless HW heater endures countless burns measuring but a few tens of seconds, 3 minutes (180seconds) is a long by comparison. With exterior controls one could increase the hysteresis if desired, but there's little point to it in your case.

Thanks again for the reply. Easy for you to say, but there will be a bit of intermission while I process it.

Do you happen to know how I might attach the Knight HW sensor (designed for their tank) to the Ergomax? That would allow me to set it through their controller and hopefully by PC if I decide to really stay on top of system performance.

I haven't seen the Knight HW sensor, but there's almost always a way. Do you have a picture of it?

Dana, I can't find a picture of the approved tank sensor -- the parts distributors don't have an image. It is part TST20015, it slides into the "tank bulbwell", and appears in a diagram in the Squire manual (http://www.lochinvar.com/_linefiles...ev%20H.pdf) on page 15. It's a small round cylinder, dimensions unspecified, attached to a terminal strip by a (long) wire.

Drainwater heat recovery sounds impressive, but for what I have would also be very invasive: Next house.

While trying to process your replies, I realized there are some basics that I don't understand:

• Your calculations refer to the boiler-loop flow rate, and I don't know how one controls that -- I had thought that the HX demanded a very high flow. Is 2.5gpm sufficient, regardless of firing rate? How do you control the flow rate? If it's by using a ball-valve, that seems inefficient in that it makes the pump work harder.
• You (everyone actually) refer to a delta-T of 20F across a zone. What makes this a good value, or is it just typical? And how do you control it?

You control the delta-T across the radiation with the flow rate (pump power for the given head & ball valves for tweaking it.) The amount of power used by ECM-drive pumps is controlled by back pressure- when you restrict the flow with a ball valve it doesn't have to drive as hard to achieve a constant back pressure. (When all zone valves are closed it's usually nearly-zero power.) The delta-T across a zone will vary with the heating load and average water temperature as well- the same radiation emits more heat at high temp than at low temp, resulting in a larger delta. At higher flow the delta-T shrinks and the overall heat emitted rises, but the pumping power use also rises. There's no magic in a 20F delta, but it's a "reasonable" design parameter for a wide variety of systems, to keep pumping power low without unduly complicating the radiation design.

The flow requirement on the boiler are boiler-specific, and are also determined by pump power & the combined head of the boiler, plumbing and tank, tweakable via ball valve.

Every time I consider a change in configuration, I re-read your responses above. I was finally able to get through it and follow your reasoning. Thanks.

Now I am trying to understand how the system might be wired and set-up on a Knight boiler. I have complicated the problem by getting cold feet over having sufficient DHW. I am trying to decide between the 2 configurations and to understand how to set up the controls. I would like to get your opinion on the alternate configuration options, but especially on the ability to control it using the Knight controller. If the controller is sufficient, I could then have both console and computer access to all temperature settings.

I am being a specific as I can about how I think the system might be controlled and set-up. The only way I can figure to control the BufferTank so that its heat is used before firing the boiler is to set the SYSTEM_SENSOR to be the CONTROL_SENSOR. Will this give reasonable control of modulation, or is there a better approach? The modulation seems pretty removed from the settings I can can control.

Ergomax-centric Configuration with Knight KBN106 and Ergomax E26 (E44?)

I'm considering stepping up in size from a 80 to 105mbh to insure a 2.5gpm shower. This violates the rule to size boiler for SH and not DHW, but the buffer largely removes the consequences of violating it. Does that seem over-cautious? The control issues are the same regardless of size.

• The E26 provides HW and is piped as hydronic separator. The Ergomax has lost its edge since rapid mixing has lost its advantage.
• SH zones would be piped off a single Variable-Speed ECM Delta-T pump with zone valves
• Set DHW-TYPE=ZONE. Use a Knight DHW sensor to monitor the Ergomax temperature, hopefully in an immersion well. This should keep the HW at its set temperature when ODR calls for temperatures below DHW. I would initially try running DHW at 125F/115F.
• Set CONTROL_SENSOR=SYSTEM_SENSOR, and put that sensor on the system loop. System, and hence the zones, should track ODR parameters when it calls for temperatures above the DHW setting, but cannot go below the DHW/buffer setting. I would initially try running System at 135F/125F, with ODR set at 145F at 10F, and 100F at 50F.
• My calculations show that it would take 6 minutes at the minimum 20mbh firing to raise the E26 buffer 10F. The worst-case cycling would be when half of that output goes to SH heating & half to the tank. yielding 12.5 minutes on and 12.5 minutes off for a resultant 2.4 cycles per hour. Would that be reasonable? If I upgraded to the E44, it would take 11 minutes to recharge and have a worst-case 1.3 cycles per hour.
• The DHW would call for heat after about a 4 gallon draw with the E26 (8 gallon with the E44), and would have to be recharged at a high firing to insure hot water.

Traditional Indirect plus Buffer with Knight KBN086, 40-45 gallon Indirect and 26 gallon BufferTank

I wanted to explore a system optimized for performance, while giving the expected long shower or tub-fill. This configuration adds an Indirect, which sits across the primary, but keeps a 80mbh boiler and can use a simple buffer-tank. It suffers the standby loss of 2 tanks, but can track the lowest viable SH temperatures. Does this make sense for baseboard heat?

• The buffer tank is piped as hydronic separator.
• SH zones would be piped off a single Variable-Speed ECM Delta-T pump with zone valves
• Set DHW-TYPE to NORMAL and use the usual Boiler & DHW piping off the primary. The DHW pump should not run frequently, so it could be non-ECM.
• Squire SIT40 might be appropriate for the Indirect because it can accept the Knight DHW temperature sensor. Set DHW to 130F/115F if that will satisfy our needs, bumping up if necessary. I want the DHW to recharge at low fire: Would a boiler set point of 135F/125F accomplish this?
• Set the zones using ODR and the lowest viable temperatures. I would initially try 145F at 10F, and 100F at 50F. Your suggested 150F Boiler Outlet temperature seems reasonable, but how do you decide between higher temperature vs higher flow?
• Use the SYSTEM_SENSOR, set as CONTROL_SENSOR, to control the SH, with the sensor placed in the the secondary loop.
• My calculations show that it would take 8 minutes at the lowest 15mbh firing to raise the buffer 10F. The worst-case cycling would be when half of that output goes to SH heating & half to the tank, yielding 16 minutes on and 16 minutes off for a resultant 1.8 cycles per hour.