High efficiency water heater vs wall mounted modulating boiler I know it's been discussed to death but I need a little help and am hoping someone can give a hand. Dana1 are you out there?

We're replacing the heat source for a roughly 4500 sf eight year old home. The envelope is very high performance. In addition to being well insulated it was constructed to be exceptionally air tight. The crawl space is unvented, so effectively it's a very short basement. The windows are top end double pane, but not triple (too expensive at the time).

The current heat sources are two 5 ton air sourced heat pumps with reverse cycle chillers. There are two 100 gallon tanks, one for hot, one for cold water. Hot water is used for the radiant heat, the cold to feed two air handlers for air conditioning. The original idea was that Portland Oregon is temperate with very low cost electricity and at the time the building was being designed the thinking was that natural gas would go up from significantly from the then cost of $2 per therm. Instead the price of gas went down and to encourage energy efficiency Oregon increased the tax on electricity.

In addition there have been problems with the current equipment. The heat pumps shut themselves off at low temperatures and the reason remains unresolved. Also the heat capacity of heat pumps falls off dramatically as the weather gets colder such that at 9 F that 120k BTU of capacity is down to less than 60k BTU. So just when you need the most heat you get the least. Likewise efficiency of the heat pumps falls off so you end up with not enough heat to meet set point on the coldest days but paying more for the privilege of being cold.

Finally there was a communications problem between the guy who specified the design for the radiant tubing and the ones who installed the heat sources. The floor uses omega plates on standard spacing, standard that is for water coming from a gas fired appliance. The heat pumps aren't thrilled to be asked to pump out 120 F water and this has resulted in a number of issues.

Bottom line it seems like the right decision is to leave the heat pumps in place for air conditioning but use the space in the equipment room, which is ample and fully within the envelope of the house, currently occupied by the domestic hot water heat to put in a gas heat source.

One option I considered was a modulating boiler, such as one of the better wall mounted units. A few have dual piping making it easy to connect up to an indirect tank for DHW. But the vendors tell me that these only perform well if the delta T across the loops is high, ideally 30 F. With tubing on perhaps two foot centers (I can’t remember exactly) and the standard omega plates looking at the pictures I took before we installed the hardwood flooring I’d estimate the coverage for the aluminum is less than 50%. With 3/4” of wood over that kind of setup the thermal resistance between the water and the room isn’t fantastic. Under low load conditions its not clear to me how to achieve the desired delta T and therefore a low enough returning water temperature so that the boiler would operate efficiently. On the other hand I don’t believe the tubing arrangement that was used is wildly out of the ordinary. So my first questions.

Was the design for the tubing adequate?
If so, how does one assure that the returning water temperature will be low enough to keep a wall mounted type boiler in condensing mode?

Isn’t it true that under low load conditions no matter how low the thermal resistance is between the tubing and the air in the room the delta T across the loops will be low so the returning water temperature will be high?
If so then how is it appropriate to ever use a heat source like these?

Alternately I looked at the high efficiency water heaters. These don’t appear to be affected by the returning water temperature. If that’s the case are they always the better choice?

Sorry for the long post but I wanted to get all the information out there. Thanks to anyone who can help me understand this better.

You can enter breaks with "<br>".

Heat pump systems are notorious for having relatively high failure rates and quirky behavior. Boilers, especially electric, are far more robust in that regard, but provide only about 1/3 the coefficient of performance. In all cases, HR systems need to be properly designed and the heat source needs to be properly selected and properly installed.

Low temperature systems such as slab-type HR floor heating, dedicated domestic water heating, swimming pool heating and snow melting provide condensing gas boilers with low inlet water temperatures allowing them to achieve condensation and high efficiency over most operating time. HR systems that are designed around water temperatures higher than about 140F may not provide boiler inlet temperatures low enough to achieve condensation and higher efficiency during much of the operating time.

Not sure the line of thought on a 2 foot spacing when using Omega heat plates, We design these systems daily and usually will run an 8-12 inch pattern.
30 degree delta T is high, typical is more like 20.
Last are you describing a sandwich system, on top of sub floor or are you a staple up, under sub floor. Not clear in that you suggest 3/4" on top of plates.
That said a typical sandwich system like our RHT floor panel system ( plates on top of sub floor) can easily drive well insulated homes in the 100 degree range.
If you are running a staple up (plates under sub floor) and it is working well at 120 degrees that's great.
To the boiler, Modern condensing boilers will turn down 5 to one and some even 10 to one, 100,000 BTU can drop output on 5/1 turndown to 20,000 BTU.
On a 4,000 square foot house with a 20 degree sq ft BTU load you are looking at a fairly small unit, often if there is a side arm tank the boiler is sized for the DHW load over the heat load.
So all that said, a condensing boiler these days might be a good choice,
If you can design to keep the heat pump in place and be supplemental all the better.
Dan

The one part of my long post that I wasn't sure on was the spacing on of the tubes. I went back and yes it's one foot. Sorry about being vague regarding the piping installation. It's is in fact a sandwich. Floor joists, then a layer of plywood, then a second layer of plywood strips with slots for the PEX and omega plates, including slots at the ends of each loop for turn arounds. The final layer is the wood flooring. So there is only one layer of wood above the tubing and plates.

I've attached a picture of my layout. Assuming I used 5" wide plates, which sounds about right, then along the straight parts of the tube runs about 40% of the floor is covered in aluminum plates. Allowing for turn arounds it would be a bit less overall.

Because the plates are only where the runs are straight and because of the width of the plates I think the coverage is about 50% but I'll go back and pull the photographs to confirm this and post the result.

My concern is that when I've used a very high accuracy temperature problem with a flat head designed for measuring the temperature of a surface to check the source and return temperatures of the manifolds the delta T is no where near even 20 F. Maybe I'm doing something wrong but I figured that the temperature of the brass manifold is pretty darn close to that of the water flowing through it since brass is pretty conductive. This fact alone has me worried. If the thermal resistance between the water and the room is too high (perhaps because the plate coverage was too low) then that would certainly result in a low delta T.

So I guess that's the place to start. Shouldn't my first step be to confirm that the loops are either performing reasonably or that they aren't?

The system used electronic actuators for every loop. Not every room, every loop. Rooms with more than one loop just have the actuators wired together and are controlled by a single thermostat. One thermostat per room. There is a single, smart ECM pump. It can be set for constant delta P (as I remember) so it controls its speed adjust flow as zones open and close. Each loop also has a flow meter so I can see that the rates are reasonable, between 3/4 and 1 rpm. The only remaining variable I think is the delta T and the way I measure it I'm getting just a few degrees which is either wrong or extremely problematic.

Reducing the flow rate will increase delta T. If a room has multiple loops, than each loop should have an actuator. The actuator controller is wired to control all the actuators associated with the thermostat zone. So that is okay. Most manifolds have supply/return temp gauges. Just measure your temps on the supply/return lines near the heat source.

I wouldn't get too hung up on the delta-T, especially at this time of year (low loads). The efficiency difference is a few percent (ie, they still perform well) and 30F variation in radiator output may not provide a comfortable floor.

With an above-floor emitter and Portland OR climate, a mod con gas boiler should operate well. Still, it is always wise to run the numbers and get minimal size boiler that satisfies an accurate heat loss analysis.

Posted By jonr on 15 Apr 2017 10:18 PM
I wouldn't get too hung up on the delta-T, especially at this time of year (low loads). The efficiency difference is a few percent (ie, they still perform well) and 30F variation in radiator output may not provide a comfortable floor.

So here is my question.

A low loss structure combined with Portland's mild wether I can easily imagine that the result would be that a condensing boiler would almost never be condensing. Am I right? If so then I don't see how using one makes much sense.

I don't want to get hung up unnecessarily on the delta T and the resulting returning water temperature but I'm thinking that this is at the core of the decision regarding of what kind of heat source to buy.

But perhaps my thinking is just way off. Any help setting me straight would be appreciated.

Condensing boilers start condensing any time boiler set point temp is below 135 degrees.
The colder you set it say 90 degrees for your sandwich system the greater the fuel efficiency.
Modulation, turn down and out door reset further add to the ability to add efficiency.
Dan

Posted By Blueridgecompany.com on 16 Apr 2017 01:02 AM
Condensing boilers start condensing any time boiler set point temp is below 135 degrees.
The colder you set it say 90 degrees for your sandwich system the greater the fuel efficiency.
Modulation, turn down and out door reset further add to the ability to add efficiency.
Dan

Ah ha. I was focused only on the problem of short cycling as reason for outdoor reset because that’s a big deal for heat pumps and hadn’t though about how decreasing water temperature would help efficiency.

Question. I know that the water temperature is often/usually set back by sensing the outdoor temperature. This has never made sense to me since that isn't necessarily a good indication of the load. For example it can be cold and sunny in which case our house can actually go negative i.e. it warms up beyond the set point solely due to solar gain. So why not set the water temperature based on the duty cycle for the call for heat or the flow rate of the variable speed pump? These are direct measurement of the load.

So for example one could look at the duty cycle of the zone control valves and vary the temperature to keep them at some preset duty cycle, say 75%. 100% would be better but 75% (with no zone running at 100%) would allow for room to room variations. Alternately one could read the flow rate if one was using a variable speed pump and set the water temperature to whatever rate was optimal (a trade off between efficiency at the boiler and pumping losses).

Does anyone do this? If not any idea why not?

Yes, although in your example of negative load with OR, all zone valves will close and the boiler won't deliver any heat, which is as efficient as you can get. You will probably enjoy reading these:

http://www.tekmarcontrols.com/images/_literature/e0004_06.pdf?lbisphpreq=1
http://www.pmengineer.com/articles/86699-outdoor-reset-control-no-quality-hydronic-system-should-be-without-it

Also consider buying his book.

Simply stated, heat source minimum operating time (and potential for short cycling) is a function of the heat source Btu output and the volume of hydronic fluid in the system and actual loops receiving the heat. This is why modulating boilers and buffer tanks may be used to prevent heat source short cycling. We have a calculator on our website to calculate heat source minimum operating time.

Simply stated, outdoor reset varies the boiler supply setpoint temperature depending on the outdoor temperature. When the outdoor temperature is colder, the boiler supply setpoint temperature becomes higher. When the outdoor temperature is hotter, the boiler supply setpoint temperature becomes lower. This effectively modulates the heating rate of the HR emitter based on the actual heat loss rate of the building which is largely a function of the outdoor temperature and significantly improves the controllability of the HR system. The best book on HR systems is “Modern Hydronic Heating” by John Siegenthaler, PE.

Deleted

Dan, please be advised that there is another post and person specifically asking about your plates. It's a very hard post to read as there aren't any sentence or paragraph breaks...

Posted By sbe53 on 16 Apr 2017 04:44 PM
I don't see any diagrams showing that buffer tanks are needed?

Buffer tanks are only needed when they are actually needed to prevent short cycling. One needs to actually run the design numbers to see if needed...or actually experience heat source short cycling...which is especially rough on heat pump systems and gas boiler heat sources. Otherwise, buffer tanks are best avoided.