New construction hydro-air vs forced air with high-demand shower

[aardwolf]

I'm building an 8000 sqft (including basement) house outside of Boston. We're trying to choose between forced air and hydro air.

I can't find that much out there about how to power a high-demand shower. We are installing a custom shower with 2.5 GPM showerhead, 2.5 GPM handshower and 3 x 2 GPM body sprays, which can all be run simultaneously (11 GPM). We also have a tub that's 84 gallons to the overflow, but probably will fill it to around 50 gallons. The tub filler is ~14 GPM.

The two proposals are:

(1) forced air and tankless hot water:

5 furnaces (!!!): 1 handling the finished basement, 2 handling the 1st floor, and 2 handling the 2nd floor and not-yet-finished attic. 2 could've handled the downstairs but making it 3 helps with duct constraints.
3 Navien HWHs: 2 cascaded to handle all the upstairs bathrooms including that master shower, and 1 to handle the basement and 1st floor (I think we could do 2 for everything)

(2) boiler + ~80gal indirect or a TurboMax:

still trying to flesh this out. venting and gas lines for 5 furnaces is ugly and expensive which prompted looking into a boiler. I'm not super worried about the HVAC side of that. we'd probably still have 5 air handlers because of ducting.

I've read that tankless hot water is better for big showers. if we go hydro, I feel like we'd need an oversized system to handle that peak load, which would be inefficient most of the time.

the turbomax seems like an interesting option. if i understand it, it's the boiler equivalent of a tankless heater. the HVAC supplier suggested a TurboMax 65 with a ~500k BTU output boiler to do 15GPM @ 110F which seems in line with 2-3 200k BTU Navien 240A's.

But can't a normal indirect heater like the SuperStor handle this? My novice calculations:

11GPM @ 80% hot water = 8.8GPM
80 gal SuperStor SSU-80 = 56 gal usable
SSU-80 first hour recovery @ >= 212k BTU = 330 GPH
Continuous flow = First-hour - (0.75 * tank capacity) = 330 - 0.75 * 80 = 270 GPH = 4.5 GPM

So I *think* this means (and this is where I'd love confirmation/correction) that the shower will drain the tank at 8.8 - 4.5 = 4.3 GPM. That gives us 13 minutes. The 56 gallons is enough to fill the tub.

A more realistic scenario is that we run the body sprays + only one showerhead, 8.5 GPM * 80% = 6.8 GPM, and it lasts 24 minutes (56 gal / (6.8 - 4.5 GPM)). I think 24 minutes is plenty for us both to shower back to back.

I think we could add in a drainwater heat exchanger like the Power-Pipe (confirmed that it's approved in Mass.) to extend it further.

The TurboMax at a similar 250k BTU can do 429 GPH @ 110F, 7GPM. That's not quite enough.

It sounds like the options are a *huge* on demand heater like the TurboMax @ 500k BTUs / 3 Naviens, or a simple traditional tank. What's wrong with my calculations that would justify anything other than the indirect tank and a ~250k BTU boiler?

Thanks!

 

[aardwolf]

sorry, i think i should've posted this in the water heater forum. started out asking about boilers vs furnaces and turned into mostly a question about DHW!

 

[Sponsor]

 

[aardwolf]

by the way, wanted to call out that I found this old thread really helpful. https://terrylove.com/forums/index.php?threads/gas-water-heater-for-bathroom-renovation.34898/

thanks jadnashua and dana for your responses there. I hope I applied it correctly here!

 

[Dana]

A little weekend getaway spot on the North Shore mayhaps?

An 11 gpm flow is (60 minutes x 11 gallons/minute x 8.34 lbs/gallon =) ~ 5500 lbs/hour, and with a 65F rise (40F in, 105F at the showerheads) you're looking at (65F x 5500lbs/hr=) 358,000 BTU/hr of water heater output rate to cover the load in real time.

With that much shower it's worth using drainwater heat recovery to add capacity and lower the fuel use. A single 4" x 60" drainwater heat recovery unit would almost do it with a single 199KBTU tankless but 3-4 on a manifold would get you there.

Taller is better, if you have the room, but you'll still need a least three to get sufficient return efficiency, and it's easier to balance flows with the plumbing design with four.

A single 4" x 60" Powerpipe delivers better than 53% energy return @ 2.5 gpm, even more at slower flows, a bit less at higher flows. So at 11gpm with four of them reasonably balanced you'd be roughly doubling the apparent capacity, and would need only half as much burner (or tank) to support the shower. A burner output of 180,000 BTU/hr (any 199K tankless, if you like) could handle the full shower load 24/365 with that much drainwater heat recovery.

Size the boiler for the space heating load (or the hydro-air handler loads, which will be a bit higher), size the tank for the tubs, and use drainwater heat recovery to manage that gia-normous shower load. The 80 gallon TurboMax won't go very far at 11gpm on it's own, even at a storage temp of 180F, but is very useful for minimizing short-cycling on zone calls, allows you to micro- zone the house and "shares" domestic hot water and space heating loads fairly nicely.

An IRC 2015 code-minimum 8000' house would likely have a heat load of about 100,000 BTU/hr, and could be in the 75K range with attention paid to efficiency aspects during the design phase. Simulating the house in BeOpt is worthwhile, if your architects know how to use it. An iHTP Versa Hydro sized for the tub loads plus four R4-6os or taller are probably going to be a pretty good fit.

But don't commit to any of it until you've run some AGGRESSIVE Manual-J load numbers on it after doing some performance tweaks on the building envelope, using BeOpt or similar, and sizing the air handlers or radiation appropriately using 120F-140F entering water temperatures at the 99% outside design temp for your ZIP code.

 

[aardwolf]

Wow, thanks for that response!

Let me try to unpack this a little to be sure I understand.

TurboMax vs indirect tank:

I'm not so clear on the difference between a TurboMax and a normal indirect tank. I understand how they each work as heat exchangers.

• TurboMax has to keep itself full of hot boiler water, an indirect has to keep itself full of hot potable water; seems like the same BTUs either way.
  
  
• A TurboMax can run continuously by reheating the boiler water, but isn't that also how an indirect tank recovery works? Why can the TurboMax run continuously at a higher rate?
  
  
• I don't actually need to run the shower continuously! More like for 20-30 min. An indirect's stored hot water can make up for a less-than-continuous recovery rate. The TurboMax specs list first-hour rates too, so it must be able to use its stored energy in a similar way. How do I figure what size of either kind of tank I need to handle 11GPM for 30 min rather than continuously?

The TurboMax does seem to have some benefits from storing boiler water instead of potable water, but they seem maintenance-related not performance related.

When you talk about micro-zoning the house with a TurboMax, does that only work for baseboards? We're doing fan coils so I think we can only have as many zones as air handlers (plus dampers).

Power-Pipes:

I didn't know you could parallelize the power-pipes. Cool, but we're talking some $$$ for 4 powerpipes, $4k-$5k just for parts. I guess we'd make up the $4k via a cheaper boiler and lower gas costs.

Each one has a ~2 psi drop; do we lose 2psi or 8? Or is that a dumb question because water pressure at the shower determined by something else, like the water heater?

53% energy return means that if we put in 199k BTU/hr, we'd recover 105k BTU/h and it'd be like a 305k BTU/hr output. By "roughly doubling", you mean 1.53x?

 

[Dana]

Roughly doubling means you're only taking heat out of the water heater at half the rate that you would without the heat recovery units. Using your math, it's like adding a 105,000 BTU/hr burner that uses no fuel.

It's probably cheaper on a first cost to go with a big-burner big-tank Versa than to go with a 199K tankless, 4x PowerPipes, and few furnaces. The PHE199-119 has 120 gallons of storage, and a 199K burner, and the space heating side can deliver 135,000 BTU/hr of 160F water when it needs to. At 11gpm and a 140F storage temp the buffering of the tank would give you at least 15 minutes of 11 gpm shower before it started to run out even with the burner off (but it'll be on), and with 199K of burner the recovery time is fast. If your space heating load is drawing 100K of the burner output, you still have 99K of "extra" burner, and the recovery time on the tank is still faster than a typical 50 gallon standalone.

But the big burner is still burning gas, unlike a PowerPipe.

Run the Manual-Js, room by room, zone by zone, and resist the urge to oversize the air handler coils by more than 1.2x (1.4x absolute max). You'd be able to fine tune the output downward with lower water temps after the fact, but getting it about right initially makes it easier. If the air handlers are size 2x for a 100K load, when all zones are calling for heat there's nothing left over for heating the tank back up.

The primary advantage to a TurboMax relative to a standard indirect is that no how tiny the heating zone, the boiler won't short cycle. That's about the extent of it. The down side of a TurboMax relative to a standard indirect is there's no such thing as giving priority to the hot water call. I have a similar tank on my system (ErgoMax 48), and I set the aquastat on the one air handler zone to turn off when the water temp drops to 110F, to save the person in the shower, which is a "sort of" domestic hot water priority. (At 120F the air handler is pulling ~40,000 BTU/hr out of the tank.) With the air handler off I have no problem keeping up with a 2.5-4gpm shower, But with an 11 gpm shower and air handlers that can draw the full boiler output you're kind of stuck.

 

[aardwolf]

Found this in the PHE199 manual: "MASSACHUSETTS STATE PLUMBING CODE REQUIRES A DISTANCE NO GREATER THAN 50 FEET FROM THE APPLIANCE TO THE FAN COIL IN THE AIR HANDLER."

What? That's not far enough to reach air handlers in the attic (3rd floor) from the basement.

Edit: it's under "APPLIES TO DOMESTIC WATER OUTLET CONNECTIONS ONLY" which I believe is referring to the auxiliary connections. Elsewhere in the manual it says "NOTE: Do not connect hydronic heating module to air handler units. This module may not be applied to air handler applications." I'm not clear whether it's saying the aux is the only way to connect air handlers to this unit, or if it's just saying that *if* you use the aux connection, it's potable water and this restriction applies.

 

[Dana]

Alas, I think you're reading the manual correctly! :-( (I've never seen a Versa used with hydro-air, but didn't think it would be a problem.) The distance is only an health issue if it's potable water in the hydro-air loop, but there must be something in the control algorithms on the isolated heating side that makes the Versa less suitable. Good catch!

There is probably a TurboMax + ~180-250K boiler (with or without PowerPipes) that works. It's a bit silly to be specifying the equipment in this much detail (let alone soliciting bids), before the complete envelope design is known (including window glazing/sizes, all R-values, etc), sufficiently locked down that a careful Manual-J load calculation can be performed. But it's never too early to be researching the rough outlines of the approach. Don't let the HVAC contractors do the load calculations- the industry track record is abyssmal. Hire a third party such as a registered P.E., RESNET rater, or an architect who knows what they are doing run those number. Many of these third party number crunchers can do the duct design as well.

 

[aardwolf]

I think the envelope design is known, just not to me. My builder has done this floorplan a couple times (minus some changes we made, and its orientation to the sun). I haven't seen the Manual J for those builds but the HERS report is public record. Here's one, I don't really know how it compares to Manual J but it's interesting regardless. Looks like they did furnaces in that house.

 

[Dana]

To start with, right at the top under infiltration, 4 ACH/50 doesn't even meet MA code- it has to be under 3 ACH/50 for new construction. Was that an actual blower-door test?

 

[aardwolf]

Oh, great catch! Yes, reading more closely that's the estimate. Here's the final.

Latest idea being floated is two Lochinvar WHN156 (144k DOE) or KBN151 (139k DOE) boilers and a Turbomax 65. The two boilers is a new consideration for me. It's more expensive and less powerful than a single KBN400 (373k DOE) or WHN400 (380k DOE) so I'm not sure what the pros/cons are... that'll be my research project this weekend.

 

[Dana]

Get a zone by zone, room by room heat load calculation done by a competent third party, but also:

A U-0.30 window is pretty much a code min (= U0.32) window, and an SHGC of 0.30 is sub-optimally low. In this climate it's worth the up-charge for double-pane ~U0.20-U0.23 dual l0w-E window with an SHGC of 0.50 or greater. The primary difference between that and a U0.30 window is the second hard-coat low-E coating on the side of the glass facing the room, which reflects room heat back into the room, body heat back to the warm humans. It has all the comfort & efficiency of a triple pane, but not nearly as expensive. The low-E coating on the conditioned space side of the window lowers the temperature, making it more prone to condensation (which defeats the low-E of that coating), but that doesn't happen until it's in negative double-digits outside, which doesn't happen often enough to matter in a MA climate.

When there is condensation on that type of glass it's still running U0.32 (50% more heat loss than the double low-E, but still code-min performance.)

Cardinal Glass LoE 180 + i89 is one such example, used by several of the big window manufacturers, but there are others.

This will of course increase your cooling load a bit- you may want to use a lower SHGC on west facing glass to kill the late-day/low-sun gains, but on all other sides it works pretty well in this climate. It will also reduce the portion of the heat load attributable to windows by 25-30%.

[edited to add]
The condensation line for a double low-E double pane is between the green (double-glazed, low-E) and orange (double-glazed, clear glass) in the graph, but closer to the orange. Assume that when it's -10F outside and 40% interior relative humidity (on the high side of where most homes would be, in the middle of the 30-50% RH range recommended for human health & comfort) there could be some condensation. At 30% RH (still in the human-healthy & comfortable range) there won't be condensation until -15F or so. While those temperatures may happen a few times per decade in eastern MA, they don't happen every year, and when it does happen it's only for a few hours, not days.

When talking to window-nerds, you're specifying low-E coatings on surfaces #2 (the interior side of the exterior pane) and #4 (the room-facing side of the interior pane.)

Cardinal's U & SHGC charts for different dual l0w-E glazing is on pages 2 & 3. With their LoE 180 on surface #2, argon fill, the U-factor of the glass is about U0.21, and the SHGC is 0.58-0.62, depending on the spacing between panes.

Spending the money on the building envelope is a better assurance of comfort than spending it on the mechanical systems (with the exception of radiant heating & cooling, which you're NOT doing.) Windows are the weak point. With thermal bridging of the framing included an R21 2x6 wall (barely better than code-min) is about R15 "whole wall", but a U0.30-ish window is an R3. hole in that wall which reduces the mean radiant temperatures of the room by quite a bit, which requires raising the room temperature by several degrees to hit the same comfort level. Replacing it with a U0.20 triple pane makes that only an R5 hole in the R15 wall, which helps comfort by quite a bit, but even a U0.25 wall (R4) window with a infra-red reflective low-E coating on surface #4 does at least as well on a comfort level, because it reflects the 90F+ heat radiating off of human skin back to the human skin.

 

[Dana]

If you're going to divorce the water heating from the space heating (which might be the right thing to do here) a big-burner water heater with sufficient volume & burner to cover your needs is a better & cheaper solution than a separate TurboMax + boiler. A TurboMax only makes sense if you're micro-zoning, and the minimum heat emitter on the smallest zones would short-cycle the boiler.

A stainless tank version such as a Polaris, or HTP Phoenix would be where I'd look first. Both come in a number of different volumes and burner sizes (up to 199,000 BTU/hr and 120 gallons.)

This really needs to broken down by the heat load per air handler zone, and the total amount of air handler output (ideally no more than 1.4x the whole-house heat load.) That would allow you to optimally size a boiler (and potentially a buffer tank, if needed) for the space heating. Right now there are too many unknowns to be able to optimize it.

There are a growing number of fire-tube condensing boilers out there with 10:1 turn down ratios, so even if the smallest zone only needed 12,000 BTU/hr and the whole house load was 90,000 BTU/hr, a 120,000 BTU/hr boiler with a 10:1 turn down wouldn't short-cycle on just the single small zone calling for heat, eliminating the need for buffering. I would expect with five zones you would probably have multiple zones that small, or even a bit smaller. But you need to run the numbers on the zone loads as well as the total whole house load to sort this stuff out. Without it there is no way to tune or assess the proposals.