Spacepak Split Inverter Heat Pump: SIS-060A: Glycol + Head loss

rhl

Hi,

I am working with a hydronics designer on a set of plans to install the SIS-060A.

https://www.literature.mestek.com/dms/SpacePak/SIS2-0421_2.pdf

I have a couple of questions for the group to discuss/speculate on.

1. Why do they specify/require a constant 12GPM on the water side of the unit?
2. Any clue why the primary heat exchanger adds 25ft of heat at 12GPM? Is that typical of an R410<->Water HX?
3. Why is a >= 10% propelyne glycol mix required?

N.B. with a 8Cv diverter valve on 1.25" pipe to a buffer tank/reverse indirect this is ~(25+6+1.5)*1.1=35.6 ft of head for the load side of the tank. (my plans call for 38ft of head).

Re 3. from manual Pg. 8



This is somewhat mental. It appears that in cooling mode you can cause the unit to freeze the water connections !?

...its almost as if they should have a control for this (oh wait, pg 35):


In my system we oversized the cooling convectors, so the design water temp is in the 45s.

Using the grundfos 3-speed pump there pump sizer calls for ~300W when the main pump runs.

Do these (spacepak required) values seem reasonable? it seems like 12GPM at 10deg DT is sufficient for max load, but, when operating on outdoor reset, presumably we can slow the pump down somewhat at lower flow temperatures, which would drop the power consumption of that main pump?

N.B. I've tried reaching out to the presales team on the glycol bit with no real resolution, they wont comment on this, and just stand by the statements in the manual, despite it clearly being overkill. My concern here is that even the 10% derate is causing the pumps to get too big. Of course the biggest factor here is that HX. If they simply had a lower head loss on that..

Hot_water_fan

2. Any clue why the primary heat exchanger adds 25ft of heat at 12GPM? Is that typical of an R410<->Water HX?

Nope, not typical. That’s a huge energy penalty! 

hot_rod

10% glycol is a weak mix, Dow stated not belie 20\% because the inhibitors are now adequate. Unless you custom blend it

An ecm would cut the power consumption if that pump spec is correct?

hot_rod

Do you have this design manual that Siggy wrote for SpacePak. it walks you through the pump sizing. Its a couple years old so it may be an older 5 ton example. I think you can download this 3 part series a WaterWorks publication.

This graph shows a 30% glycol example for the 060 model.
25' at 12 gpm, 30%
No doubt it will take more pump than a 15-58. A Magna would be a good choice.

Your head is the highest pressure drop circuit.`

Jamie Hall

Just a couple of thoughts. First, 25 feet of head through a heat exchanger is not unheard of. It depends entirely on how the heat exchanger is designed -- and oddly, or perhaps not, more efficient and compact units tend to have more head loss than bigger sloppier ones.

There is a possibility of freezing, hence the glycol. And, just to make things more interesting, the lower the flow for a given refrigerant side power, the greater the problem is -- which makes for a nice circular firing squad, where if you get a little freezing the flow drops, the temperature drops, the freezing increases, the flow drops... and so on. I'd be extremely cautious about reducing the recommended flow below what Spacepak wants.

And last, maybe it's just me, but I've found that company -- and Mestek -- to be singularly unhelpful about engineering type questions...

hot_rod

Jim is their tech support and trainer, former contractor, he knows the product well. He has been installing A2WHP for about 20 years now. Back east!

I think Ask This Old House did an install with the unit you have, last years. Syracuse area as I recall. Probably find it in the archive.

rhl

@hot_rod your thinking of the enertech unit installed by Terry Moags group.

10% glycol is a weak mix, Dow stated not belie 20\% because the inhibitors are now adequate. Unless you custom blend it

and

There is a possibility of freezing, hence the glycol. And, just to make things more interesting, the lower the flow for a given refrigerant side power, the greater the problem is -- which makes for a nice circular firing squad, where if you get a little freezing the flow drops, the temperature drops, the freezing increases, the flow drops... and so on. I'd be extremely cautious about reducing the recommended flow below what Spacepak wants.

1. I am very interested in following the rules, unless we can prove the rules are obviously/clearly wrong.
2. I think the biggest issue is the 25Ft of head required for the 12GPM at all times. We can see that 12GPM @ 10DT produces 60K BTU of energy moved. But the unit does not always move its nominal rating..

Re glycol:

3. Can you explain to me what is supposed to freeze in this _split_ system? As i showed the manufacturer is concerned about cooling not heating.

The only legit argument i can think of is in heating mode, the R410 can still be below 32F and that could freeze the HX (say when the outdoor unit is off theoretically). But then whats the point of a split system? If you need the same antifreeze mix as a monobloc, you may as well do that, as it would be cheaper.

If there is freezing risk in cooling mode it means that at least the R16 setting shown above is non functional. That seems like something to be upfront about, given the price of the unit.

WMno57

rhl said:

But then whats the point of a split system? If you need the same antifreeze mix as a monobloc, you may as well do that, as it would be cheaper.

That is an excellent question! For an older hydronic home with no ductwork, are you stuck with glycol if you electrify? I HATE GLYCOL!
Maybe the magical heat pumps aren't so magical.

EBEBRATT-Ed

You can't fight mother nature or physics. 25 feet of head is a bit but so are most mod con boilers. If that is what the mfg says it needs that is what it needs. 12 gpm for 5 tons is about right most units need 2 gpm/ton + a glycol adjustment.

Hot_water_fan

If it’s not too late, I know the Nordic A2W has much lower head loss. It’s like Modcons, some have high head loss, some have low head loss. 

Teemok

I asked the very same question at a training class. Why such high head? The answer was a shrug. COP numbers are a little off because they don't included this necessary energy expenditure in their calcs for COP. I just installed one with a Taco 034e pump and it works very well. Note: they call for 1-1/2" piping to the buffer but the internal piping in the sim060 heat exchanger looks to be 3/4" or so, maybe less, maybe more. Didn't measure it, shakes head and closes it back up. Cheaper heat exchanger is what looks to be happening here. Externalized costs, a durable goods life time of negative efficiency ding in a green market product. Evil engineering? or just another squirrel try'n to get a nut or bean counter designers doing what they do. I love that they are making them at all, so that's the worst of my critique. Maybe a very high velocity flow is needed for peak heat transfer efficiency and the pumping cost it worth it. I doubt it. I hope they do better on the next gens. I put a delta T pump on the high mass radiant system side so water returns nice and cool but with 12 gpm stirring up the tank I'm not sure it makes that much difference. Heat pumps are hip.

rhl

@Teemok thanks for this insight. Yeah I agree with this Assesment.

In order to keep the headloss in check, in my plans I upgraded the DHW diverter valve to a 25Cv valve. The head loss (regular water) is basically 25ft + ~1.5ft for all other piping. 

Sadly the requirement to add glycol further derates things, but at least only by 10%. So similarly we are either doing a 0034e taco or a Magna1. 

However, had we not derated we cut have gone down to an alpha 26-99. It appears that would cut pumping power by 30% and cost half as much. 

Have you installed an SIS unit specifically? 

The low quality parts statement is discouraging. The unit is not inexpensive. 

Teemok

I haven't done the split but it seem like they just took the same bits as the SIM and moved them inside for the SIS. My customer just wanted radiant floor heating so that makes it easier. I heard from the trainer the 26-99 is a no go. I understand it's just a de-rating thing. Cold slab start ups would be very long. I wouldn't say parts are low quality, maybe questionable sizing design principles.