Startup of PEX radiant heat with a Rheem eco tankless heater

Users who are viewing this thread

Larry Blsair

New Member
Messages
11
Reaction score
0
Points
1
Location
Wyoming
I am starting this new system up. I have a 188k natural gas eco tankless heater and have a few questions. We live in the wilds of Wyoming. About 3 years ago they plowed a NG line in that is 250 ft from our house. The closest city to us is about 100 miles away. My wife and I built the house 10 years ago and at that time I installed 1/2 PEX in the floors. Until now we have heated with a wood stove but this year the wood cutters rebelled and said no more wood cutting ( wife and daughter ). So I have to get the radiant heat working.

I was a plumber for a couple of years so I have been able to sweat in the plumbing from and to the heater to the stainless steel manifolds ( out and in ) I am using a TACO 009 pump for circulation. The heater bottom is 4 feet from the floor. All the plumbing is below the bottom of the heater.

First the natural gas is not hooked up. The connection to the main and the meter should be installed next week. so I am doing a cold run.
I have filled the system with a garden hose and pressure tested it and everything holds pressure.
I the first question is how do you purge the air from the heater? As I added water it sounded like the cold water going in always bubbled.

I have started the pump and it sounds like it is getting an air bubble occasionally. Instead of running with a steady hummm it is passing gas every so often. I have tried to purge air at both manifolds through the petcocks there and only water comes out but I can hear bubbling in the heater. I have tried to purge air at the clean out valves on the bottom of the heater again solid water out but still bubbling sounds in the heater.
Thanks for any help and suggestions. Larry
 

Dana

In the trades
Messages
7,889
Reaction score
509
Points
113
Location
01609
Is this an "open" system, running at whatever pressure your well pump system is providing, or is it isolated from the potable supply? Most tankless heaters have problems if running at pressures under 15 psi, some have issues at 20 psi. Got a model number for that tankless?

The Taco-009 is one mother-of-a-pump. Why such a beastly pump? When over-pumped the service lifespan of a tankless will be remarkably short. Since tankless heat exchangers tolerate very high delta-Ts it's better to run the tankless at 1-3 gpm and a high delta-T. I'm getting a sinking feeling that there was very little math done on this design.

Even with manifold systems it's usually possible to purge 95% of the air out of the system by simply running a ton of fill-water through the system, injecting it at the supply side and letting it out on the return side. It's easier if it's not plumbed primary/secondary, and the boiler/tankless can be isolated from one side or the other (or both).
 

Larry Blsair

New Member
Messages
11
Reaction score
0
Points
1
Location
Wyoming
You are right no math was done on this. This is a closed loop system. I had the TACO 009 to pump through 4 4ft x 8 ft solar collectors so that is the ancestry of the pump. The radiant heat system is 10 loops interlaced throughout the house and the garage. I think there is about 2000 ft of 1/2 pex in the floor. The heater is a Rheem ECO180DVLN3-1 188k btu

I was able to get most of the air out , like you said, by hooking up a hose and running it full blast through each loop and then the heater. I am able to isolate the heater via webstone isolation valves I put on hot and cold sides on the bottom of the heater. I has have a webst0ne mixer valve so that I can keep the water temp below 85 degrees, because we have pressed bamboo flooring.

As I have been running the system the last couple of days I notice that when I turn it off for an hour or so and then flip it back on it sound like the pump starts and then it gets a chunk of air and it quits pumping. I have some flow indicators on 5 of the loops on the output side of the manifold and can see the balls go up and then drop back down when "the air hits the pump". I have the external water connected to the system so I can give it a blast of water and it seems to blow the air out of the pump and the indicators show flow.

Should the system pressure stay at or above 20 lb when the pump is off?

I can buy a smaller pump and slide it into place if it will help system life. Something on the order off a TCO 007??

Thank you for any advice you give me. I have done so many things in my life - cause I like to do it MYSELF ( my 4 year old voice) so I suspect sometimes i get in over my head and don't realize it
 

Dana

In the trades
Messages
7,889
Reaction score
509
Points
113
Location
01609
According to p.47 of the manual the minimum operating pressure at the tankless is 14 psi. If the pump is pumping toward the tankless it'll be fine to run the static pressure at that pressure when the pump isn't active. If it's pumping away from the tankless it will be lowering the water pressure at the heat exchanger, so a higher static pressure would be called for.

On the five loops with flow indicators, how much flow are you getting, cumulative? Do you have a compatible wired remote control for that unit, and if use, can it display the sensed flow rate? (I've never played around with Paloma/Rheem tankless remotes, but some tankless vendors' remotes can display gpm information.)

Without the full system schematic or knowing the actual flow rate through the tankless with the -009 there's no way to really spec the pump. Almost all tankless based systems are better off with a dedicated pump to set the flow through the tankless, and a separate pump for driving the flow through the radiation, since the radiation's flow rate requirements may exceed that of what the heat exchanger in the tankless can tolerate. You may be able to work backward from the specs in the "Recirculation" section, that indicate it needs at least 40psi pressure to deliver the maximum specified flow rate (p.47), and the "Recirculation Control" subsection on "Pump Size" on p.63 (you may have to call their help line to get real info, and they'll probably tell you that they've voided the warrantee for using it as a boiler.)

Have you done a heat load calculation on the house to even know how much heat is needed? You're not going to get a lot of heat out of it with 85F water, but an 85F slab puts out quite a bit of heat. What you'll actually need is hopefully a lot less than that, but take a stab at the load numbers using a freebie online tool to ballpark it. With tools such as LoadCalc assume the construction is totally tight, set the ventilation to zero, and disregard any infiltration losses (for now), or it'll overstate the load by quite a bit. This is for sanity-checking only- it's not a substitute for a real Manual-J that would be more suitable for designing a radiant floor, but it'll be enough to tell if you're going to miss by a mile.

Independently of what temp you're mixing down to run the slab, at a healthy (for the tankless) flow rate of 2.5 gpm (=1250lbs/hr) and a temperature rise of 40F (say, 90F return, 130F out, plumbed primary/secondary with a separate pump for the radiation) it will be putting 1250lbs/hr x 40F = 50,000 BTU/hr. If there's 2000 square feet of radiant floor that would be 25 BTU/hr per square foot, which takes a surface temp at the floor about 12-13 F warmer than the room temp. That's going require water hotter than 85F in a radiant slab with bamboo flooring on top. Hopefully you don't need anywhere near 25 BTU/per square foot. Ponder this nomograph for awhile. How thick is that flooring, and what do you figure the total R-value of the flooring & underlayment is?

Infloo2.jpg
 

Larry Blsair

New Member
Messages
11
Reaction score
0
Points
1
Location
Wyoming
Sorry for such a delay in reply. We are a commercial vineyard/winery and the last 3 weeks have been busy.

Some more information.

The entire house is insulated to R49 and is very tight. I did not want any drafts and so built it that way. We have double pane low E vinyl frame windows.

The house floor is oriented strand bamboo and appears to have a R value of less than 1. There is a 18" crawl space below that floor and a R19 reflective blanket on the ground under it. The floor consists of one layer of 1/2 OSB , then the 1/2 pex with 3/4" OSB strips between each leg of the PEX loops. There is an aluminum conductor plate that the PEX is snapped into. It goes around the PEX and then there is about a 2" wing in each side that conducts the heat up into the bamboo floor. There is approximately 1200 ft of 1/2 PEX in 6 loops for a approximate 12 gallons of volume. These 6 loops are interlaced so that the flow on one loop goes one way and the flow on the next loop goes the opposite direction to better distribute the heat. The separation between the PEX is 7". The winery has 4 separate loops, one in each 8' x 40' concrete slab. Again there is R19 reflective insulation under the concrete. Probably about 900' feet of 1/2 pex for a water volume of about 9 gallons.

I shut down all the loops except for 3 that had flow indicators on them and it appears the pump is pushing about 8 gallon per minute. I could put a valve in line and reduce that to 4 1/2 or less fairly easily or just put in a smaller pump.

I ran loadcalc and it only shows R values to 29 but it appears I only need about 13K btu. I kind of remember running a bunch of calculation when I built the house 10 years ago trying to see if I could run solar collectors and as I remember It was calling for about 85 -95k btu per day.

If I run a flow rate of 2 1/2 gallon per minute then it will take about 9 minutes to exchange the water in the flooring. I would think that the flow rate should be higher to maintain a constant temp across the total area of the floor. Maybe I am wrong about this.

If I run the heater at its lowest setting of 85 degrees what will it do to the heater in the long run?

I could build a double loop system where a loop out of the heater heats a "holding tank" to 120 degrees and then run another loop out of the holding tank with a temp mixer valve to keep at the water in the floor at 82 -84 degrees. This would allow the heater to run at higher temps and yet keep the bamboo floor at allowable temps.

Thanks for your help
 

Dana

In the trades
Messages
7,889
Reaction score
509
Points
113
Location
01609
If the only insulation under the radiant is the radiant barrier on the crawlspace floor it's potentially pretty lossy out the bottom. Even with low temp water it's usually worth installing R13-R19 fluff snugged right up to the tubing & subfloor, not a couple feet away on the floor. If the crawlspace is ultra-tight it might not be that bad.

So it's not a radiant slab, but rather bamboo flooring on a 3/4" subfloor or something? If yes you'll need to use a different nomogaph, such as the one on page 5. A 3/4" subfloor is worth about R1, the bamboo flooring is another

Again there is R19 reflective insulation under the concrete.


There is no such thing as R19 reflective insulation that takes the weight of the concrete. If it's reflective bubble-pack in contact with both the soil and the concrete it was worth MAYBE R2 before the air got squashed out of it. At this point it's just an expensive vapor barrier of effectively no thermal value. For reflective insulation to work there needs to be an air gap between it and the adjacent solid materials. I hope you installed the RB a couple inches off the gravel/soil, and poured the concrete slab with a couple of inches of air between the bottom of the slab and the RB, right? (Riiiiight...)

I suspect that in the radiant slab floors more than half the heat is going into the soil, not the house, at least until it's heated up a large amount of soil. This is the opposite of efficient. If radiant barrier is the only type of insulation between the crawlspace and the soil, the R-value of the soil is still doing more than the radiant barrier in that area but at least there's an air gap on one side- it'll give you at least R1.

I shut down all the loops except for 3 that had flow indicators on them and it appears the pump is pushing about 8 gallon per minute.

If it's pumping 8 gpm even when other branches are shut down the pump is plain crazy-oversized for the system. But shutting the other loops down increased the flow through the remaining loops. If the loops are all pretty much equal you'll get a better idea of the total flow if none are shut down, and just presume the flow in the other loops will average about the same flow as those that are metered.

If the calculated heat load is 13K, even if the crappy RB insulation is robbing performance to the point that half the heat input is lost and the true load is double that (26K), even 1 gpm (total of all loops) at 10F delta-T would be way too much. A flow of 1 gpm is 500lbs/hr, x 10F= 50,000 BTU/hr or 4x the calculated load.

With a tight R49 walls & ceiling house with only modest amounts of U0.32 window area it hardly matters how well the heat is distributed in the slab or radiant floor- the natural convection within the house will even things out substantially. But with such a low load the floor temp will be barely above room temp. With high-R houses it's often more comfortable to heat just a portion of the floor to get the temp at that section of floor up to where it delivers cushy barefoot warm-toes appeal.

How many square feet of radiant floor is there?

How much of that is slab, and how much is over a crawl space?
 

Larry Blsair

New Member
Messages
11
Reaction score
0
Points
1
Location
Wyoming
I think I am in like BOB! fired the system up today and everything works. My wife is dancing around on the floors excited about how warm they feel.
The water heater is set at 100 degrees which is probably too low. The water coming back out of the PEX at this point is about 75 degrees. I have a thermostatic mixer valve and adjusted it to put water about 83 degrees or so back into the floor. I shut the input ball valve into the heater so only allow what appears to be 4 or 5 gallon per minute per my flow indicators. It isn't short cycling yet but I could see once the system gets heated up where it might.

When I installed the bamboo flooring I had purchased much of the stuff from Blue Ridge Company and used their design feature. I put in heat defusers and then installed the bamboo over the top of that. The crawlspace is totally sealed so it is essentially dead air. I see where I should have put the r15 insulation next to the floor rather than on the ground but no changing now. This area is about 1200 square feet

In the winery part where we have a slab floor I had purchased a 4500 ( converts to about 15000 btu/hr) watt electric hot water heater a couple of years ago and it heated the area ( about 1300 square feet) and kept it at 70 degrees when it was minus 30 outside. The problem was cost of electric cause we didn't have NG yet.

Dana

The following is from another answer you posted in 2014 . Yours is in red ************************
A tankless is agnostic of all other aspects of the plumbing. All it "understands" is flow. To get it to fire up takes a minimum flow (something around 1 gpm typ.), but overpumping it will wear out the flow detectors very quickly. Plumbing the system primary/secondary with the primary (tankless) loop running something like 1.5-2.5 gpm would fire reliably without over-pumping it. A tankless heat exchanger can tolerate very high delta-Ts, so you can play around with the temperature a bit to get the firing rate reasonably dialed to the load. By separating the radiation flows from the tankless flow, the flows can be at very different rates. Pumping direct through the tankless with a fancy with a programmable pump, the flow rates will vary all over the place, and it's unlikely that it will start reliably and not over-pump, destroying the flow sensor & eroding the tankless heat exchanger from the inside out.

On low-mass burners like a tankless, always pump toward the tankless, making the cold-feed side of the tankless the highest pressure point on the system. Hang the expansion tank on the other side of the pump from the tankless to reduce cavitation issues and improve pump longevity.

To keep system like this from short-cycling it's probably better to use an electric HW tank (not wired up) as the point of hydraulic separation.

A simple control setup using an electric hot water heater as a buffer tank is to set up the zone relay with one zone to start the secondary (radiant) pump, and set up an aquastat in the electric tank as the "thermostat" on another zone to turn the primary (tankless) pump on & off. You can probably use the Taco Bumblebee as the secondary pump (even if on a single zoned system it's something of a waste.) You can either take a WAG on the amount of pumping head the tankless represents and spec a pump, or maybe Taco 0010-F3 3-speed can be used for the primary loop, setting it up at the lowest speed initially, as long as it's pumping enough to start the tankless (either way you may want to install a ball-valve to adjust the flow to a non-destructive range.) You will have to run the tankless 20-70 F hotter than you set the aquastat on the tank, so there are limitations to how high the radiation loop temps will be.

Would you explain this in a little more detail? I have the 40 gallon heater and I don't understand what you are saying about using the tank with a primary and secondary system. There is only one in and one out on the tank. I just am missing something probably very simple. I can see where this would allow a greater delta t for the heater and then allow the radiant side to run at a cooler temp.

Thank you very much for your previous comments.
Larry
 

Dana

In the trades
Messages
7,889
Reaction score
509
Points
113
Location
01609
Primary/secondary systems use separate pumps for driving the flow through the boiler/tankless and the flow through the radiation. Where the loops intersect is referred to as the point of "hydraulic separation", which is where the boiler output is transfered at one flow rate to the radiation water which is flowing at a different rate. This is often done with closely spaced tees, but can also be done with a big fat tank behaving as a tee.

79046.jpg


^^ 2 zones, with a hydraulic separation tees on each zone, one with a thermostatic mixer^^


PME_0907_Feat2Fig10Lg.jpg


^^gazillion micro-zones, but thermally massive hydraulic separator^^

A common way to convert an electric tank water heater into thermally massive hydraulic separator is to use the drain port at the bottom and the hot output at the top for the "secondary" (radiation loop) connections, with the return from radiation coming in at the bottom, the supply at the hot-out, then use the side ports for the top & bottom heater elements for hooking up the primary (boiler loop) connections, with the tankless output injected at the top element port, drawing from the bottom element. The pumping head for the primary loop is then miniscule- it doesn't take much pump, and you can use whatever monster-pump is needed for driving the radiation flows without prematurely wearing out the tankless.

With that approach to hydraulic separation the delta-T on the tankless can be set quite high with a very low gpm, without it ending up sending super-heated water out to the low-temp radiant floors. Even with the tankless set to 140F, the mixing in the tank lowers the average temp of the water at the top of the tank headed out to radiation, but the bottom of the tank will only be slightly warmer than the return water from radiation, which could be 75-90F for a low-temp system. With that approach the condensing efficiency can still be quite high despite the much higher output temp of the tankless, and with 40 gallons ( ~333 lbs ) of water in the tank it also goes a long way toward suppressing short cycling.

If you're sending out 100F water and getting 75F return from radiation that's an adequate delta-T for the tankless, but the flow might be on the low side for the radiation (or not), and yet may be too high for the tankless to handle long-term. A 25F delta-T at 6 gpm (~75000 BTU/hr) is the same amount of heat as a 50F delta-T at 3 gpm, but a 3-gpm flow is going to be a lot easier on the tankless. If you separate the flows you can set a different delta-T on the tankless, and pump like hell on the radiation if that's what it takes.
 

Larry Blsair

New Member
Messages
11
Reaction score
0
Points
1
Location
Wyoming
I have plumbed the system like your recomendation. I bought a low flow pump to circulate through the heater. I think I need an aqua stat to turn that pump off when the floor recirculating system shuts off. The aqua stat would shut the heater circulation system off at 140 degrees or what ever I set it to. I am having a problem figuring out where to put the probe. . I think the bottom side port so that when the floor circulation stops the heater would heat the tank to 140 and then turn off the heater pump.
Thanks again for your help
Larry
 

Dana

In the trades
Messages
7,889
Reaction score
509
Points
113
Location
01609
I have plumbed the system like your recomendation. I bought a low flow pump to circulate through the heater. I think I need an aqua stat to turn that pump off when the floor recirculating system shuts off. The aqua stat would shut the heater circulation system off at 140 degrees or what ever I set it to. I am having a problem figuring out where to put the probe. . I think the bottom side port so that when the floor circulation stops the heater would heat the tank to 140 and then turn off the heater pump.
Thanks again for your help
Larry

What are you using for a buffer tank- an electric water heater? An electric water heater's element switching controls can usually handle the load of a small circulation pump, so it's built-in aquastat- instead of turning on a heater element it's turning on the pump. You will probably have to make some modifications, since the electric water heaters start by turning on the top element, but switch over to the lower element for the finish.

With a well insulated buffer tank such as an electric water heater it doesn't need to turn off when ever the radiation circulation pump stops- the standby losses of a 140F electric water heater are quite small, as long as the connecting plumbing is insulated to R3 or better.
 

Larry Blsair

New Member
Messages
11
Reaction score
0
Points
1
Location
Wyoming
I have the system set up pretty much like Dana described with a 40 gal mixer tank. Its been running for about 6 weeks and really nice heat. Got the first gas bill and the weather was fairly cold, about 10 days of 10-20 degree weather, and the bill was $75.00. $2.50 a day for gas is ok!

I am having a problem with the Grundfos pump though. It just quits turning. I take the head off and turn the electric back on and the impeller just barely turns. I give it a turn by hand and it is very stiff. the motor is hot also which indicates, I guess, that it has stalled. After turning it by hand a couple of times I turn the electric back on and it spins up.


Do you think this is a bad pump out of the box? Still looking on Google for possible solutions.

I'll give a complete write up about the system in a couple of weeks so you all will see just what I did. Kind of a final report on my project :)
 

Larry Blsair

New Member
Messages
11
Reaction score
0
Points
1
Location
Wyoming
I have this pump on low speed and have closed the ball valve going into the water heater to limit the gpm to about 4 gpm. Would this cause a "dead head" situation on the pump and cause it to heat up to where it won't turn? I really don't know what dead head means so am guessing.

The system is running about 18psi.

The pump is located on the floor at the bottom of the 40 gal mixing tank so there is no air in the pump. The mixing tank is about 44 inches tall so water would be supplied to the pump at all times.

The pump is running with the shaft in a horizontal position

The plumbing in the system is all 3/4 inch copper

There is no scale building up on the pump.

I took a chance on the Grundfos pump cause I have always used TACO before. I am getting a return for the Grundfos and replacing it with a TACO but have to keep the system limping a long for the next week until the new pump gets here. It's about 18 degrees out so I have to have heat ;)

Thanks for any ideas.
 

Dana

In the trades
Messages
7,889
Reaction score
509
Points
113
Location
01609
Which model Grundfos or Taco are you using for a primary pump?

Pumps do sometimes fail right out of the box, but it's rare. It sounds like the bearings have failed, but it could also be the motor itself. A very high-flow higher HP pump might be working extra hard at 4 gpm, but most pumps are fine with that.
 
Top
Hey, wait a minute.

This is awkward, but...

It looks like you're using an ad blocker. We get it, but (1) terrylove.com can't live without ads, and (2) ad blockers can cause issues with videos and comments. If you'd like to support the site, please allow ads.

If any particular ad is your REASON for blocking ads, please let us know. We might be able to do something about it. Thanks.
I've Disabled AdBlock    No Thanks