CH System Overhaul with Cast Iron to PEX Conversion
Hello everyone. I am planning to upgrade my CH system and would really appreciate some input.
Current System:
- Two pipe, direct return. 140MBtu Dunkirk boiler serving 8 CI radiators of different shapes and sizes. 82% efficient boiler per nameplate and combustion analysis.
- Separate, radiant floor system. Takagi tankless "boiler" (I know...) serving two zones of heated floors. I built this as a prototype when renovating/eliminating some CI rads and it has worked well for around 4 years.
Home Details: Three story 1925 construction (brick and block exterior) with moderate insulation added over the years. 2400 sq ft. Pittsburgh area. Detailed heat loss calc results in 114MBtu (68F/0F design criteria, including piping loss).
Objective: Ultimately, replace both boilers with one high efficiency boiler to serve the high and low temp zones. Run PEX home runs from manifolds to existing radiant floor loops and CI radiators (not modifying any). Add three panel radiators to the low temp zone. Eliminate the existing two-pipe system piping in the basement (which runs the perimeter of the basement and is a real head banger in most spots). This will result in a significant amount of space savings in the basement, correct several design flaws with the existing CI system, reduce maintenance, and hopefully gain some efficiency.
I attached a schematic for the planned boiler piping. I wanted to point out Zone 4, which is a very small load. However, in >4 years of operating, it never calls for heat on its own and is always in tandem with Zone 3 (which is the kitchen/pantry/hallway with open spaces and lots of windows). Hopefully this is not an issue with the new design.
Do you see any major concerns or flaws with what I am planning? Also do you have any recommendations for boilers to suit this application?
Thanks in advance!
Al
Comments
-
Forgot to mention on the schematic, thicker lines (supply/return for CI) is 1-1/4", thinner lines (low temp zones) is 1". 3/4" PEX to all CI rads, 1/2" PEX to all low temp emitters/floor loops.
0 -
A mod con is probably a better boiler for that system, you can run odr on it set for the ci radiators then mix it down for the other loads but the mass of the system is likely something you'd need some protection from low return water temps on a ci boiler. Might want to use a control to do the low temp zones that can do odr too.
Do you really need 170 to the CI ever? It is possible you might but is likely you don't.
2 -
First step is to get better handle on your actual heat loss. I know you said "Detailed heat loss calc results in 114MBtu" but that is unlikely for a place that big.
If you have last years winter fuel bills, a rough heat size is to convert the fuel used to therms and multiply by 42. (ie to have a 114m heat loss you would have used about 114000/42=~2700 therms).
Once you have a better handle on heat load, you can work on sizing boiler and required system temps.
With a mix of high temp and low temp, I would do a PS loop instead with closed spaced T for each temp zone. Put the high temp zones pulling off the loop first followed by the low temp zones. This way the low temp zones will be partially fed by the return water from the rads. Increases loop delta T which is better for efficiency.
0 -
Some of the mod cons have multiple temperature outputs. But only one runs at a time. Maybe the high mass warms and can be off while the boiler goes to the low temperature load?
Not many series primary loops being used these days. If the high temperature load is not running, the low temperature will not see that lower return. Plus you need a loop pump if the boiler needs hydraulic separation
Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream
2 -
If you pick a boiler with multiple DH calls with different setpoints or reset curves you can make the boiler only heat to the current temp needed but you will still need a control to mix down the hotter water when there is a simultaneous call for the higher temp zone.
0 -
I thought they can only do one temperature at a time. So it could not provide the lower temperature load while the high temperature is calling?
Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream0 -
It can only do one temp at a time but when only the lower temp zone is calling it can use that as the setpoint instead of using the higher setpoint then having to mix that water down to the lower temp.
It should be a little more efficient when only the low temp zone is calling but it still needs a thermostatic mixing valve or injection mixing for when there is a call for the high temp zone.
0 -
I was not talking about a series loop. Standard primary secondary piping with the high temp zone pulling from the first set of closed spaced Ts, the low temp from the 2nd set of closed spaced Ts. You would still need a mixing valve for low temp and a pump for primary(if not inside boiler) and for each secondary temp zone.
0 -
I believe Taco makes a ODR mixing valve! You set the ODR from the boiler to operate the high temperature zone and you set the Taco mixing valve ODR for the low temperature zone. Basically you’re resetting the low temperature zone off of the already reset high temperature zone.
Edward Young Retired
After you make that expensive repair and you still have the same problem, What will you check next?
1 -
I can’t speak for every brand, but the ones that I’m familiar with will run both zones if they’re calling, but default to the higher heating curve. Obviously therefore, a mixing device is needed to limit the SWT to the lower zone when both are calling.
Bob Boan
You can choose to do what you want, but you cannot choose the consequences.2 -
UM, how about giving us a schematic of the heating system as is??
0 -
Definitely agree with mod con and ODR. No, I don’t need 170F often – maybe if the winter is brutal, which we encountered in mid January and the CI boiler hit limit (175F setpoint) but there was still a call. Point and click thermometer showed ~150F/~135F (supply/return) across most of the CI rads. Those days are very few and far between though. I ran some more calculations and the vast majority of the heating season can be satisfied with an ODR curve with less than 170F supply temp.
0 -
You are right, I had a lot of conservatism baked into the heat loss calcs. I did a room-by-room U*Temp Difference analysis including infiltration (28% of room heat load) and included floor perimeter losses, following a design manual by Vanguard Pipes & Fittings. Since the original post, I removed floor perimeter losses and set the design outdoor temp to 5F (per 2018 IPC) instead of 0F. I kept infiltration losses at 28% of room heat loss though. Revised number is 82,000 Btu for everything on a 5F day. I used the USBoiler.com heat loss calculator as a quick check which gives 78,000 Btu (but doesn’t seem to specifically include infiltration or pipe losses).
Regarding the fuel bills - Average therm usage over 3 years of heating seasons is 1160 therm. 1160*42=48,720 which is much lower than my estimated loss. If my heat loss calcs seem inaccurate based on this, I am still open to any input.
Pulling off the return of high temp zones to mix down to the low temp zones would not work in my case, because the low temp zones (staple up radiant, etc.) call much more often and for longer durations than the high temp zones. The heat from the high temp zone returns just wouldn’t be there in most cases.
0 -
Thanks for the input!
I am certainly thinking a mod-con boiler with ODR on the new system. ODR on the mixing valve will be considered in the future to keep it simple for now.
I attached schematics of the existing two systems and the proposed new design with revised heat loads. My biggest concern at this point is the small Btu load of the low temp zones. The tankless water heater has been chugging along, heating the low-load radiant floors for years, but I want to make sure the new boiler can do the same without any negative effects.
0 -
If you pick a boiler with a 10:1 turndown ratio with about a 80,000 btu/hr output and some ways to set it to ramp up slowly instead of trying to hit setpoint immediately, the 4,000 btu/hr load should be the only one that is a concern. If there is a lot of volume of water in the tubing that may be less of a concern. It will cycle but the volume of water will tell you how long those cycles are.
0 -
That 4,500 Btu zone is only about 6.8 gallons of water (about 700ft of 1/2 PEX tubing total). It never calls for heat on its own though. If it is calling, the other low temp zone is already calling, so it is always running in tandem with the 12,500 Btu load. Unless of course the thermostat was cranked up on that zone, but that is unlikely.
1 -
of course keep in mind if you are using a lower water temp there is a lower output so odr with those zones could cause it to cycle.
0 -
Since the common small boiler is 80k, it won't make much difference if your heat loss is 50k or 60k. My guess it is somewhere between the on-line calculator and the rough fuel use estimate. If you want a more accurate number, you can run through math here:
If your main heat source is the floor heat, I would try get a modcon that can do variable supply temp. These will keep the supply water low when only the floor heat calls and only increase up to the radiator temp when those zones also need to run.
I don't think you need the swing check on the floor heat return.
0 -
If you use a common 3 Cv thermostatic mixing valve, 6 loops is about max.
Either two mix valves one for each manifold, or a valve with a 5 Cv.Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream0 -
That's really good info. Using the method in the GBA article (using fuel usage to determine building heat load) I get 52,000 Btu. That's based on 3 years of bills and heating deg-day data, and I subtracted out therms used by the water heater and gas stove, so I feel good about that number.
The floors only account for ~1/5 of the entire heat load. The bulk is done by the CI rads throughout the home.
Roger that on the swing check.
0 -
I certainly value your input here, but I wanted to make sure I understand. The current system delivers ~3 gpm total to the 7 existing loops (as indicated by the screen on the tankless heater). I plan to add one 4 more loops (one staple-up and 3 panel rads) which would increase flow needed to ~5 gpm. Is that too much to pull through a 3 Cv (planning for 1" piping and 1" x 1/2" manifolds)? If that isn't advised, can you recommend a 5 Cv valve?
0 -
Cv is the flow through the valve in gpm at 1 psi pressure drop across the valve. It should roughly match the flow through the valve. Taco heat motor valves have higher Cv. I think some of the Honeywell motoraized valve do too. Look at at White Rodgers and Caleffi too.
0 -
I saw the 12 port manifold and thought maybe you were planning 12 loops at .65 gpm each, 7.8 gpm
If so here is what that looks like form pressure drop, third cell.
5 gpm is workable with a 3 Cv valve. I attached that example also
,
Bob "hot rod" Rohr
trainer for Caleffi NA
Living the hydronic dream0 -
If you are close, CV is less of an issue if you have a pump that does a bit more pressure. Some of the ECM units are rated at 20' much more than your typical 007. You still have to watch that the pressure before the pump doesn't drop too much, you always want about 5psi there to avoid cavitation.
0
Categories
- All Categories
- 86.8K THE MAIN WALL
- 3.1K A-C, Heat Pumps & Refrigeration
- 56 Biomass
- 425 Carbon Monoxide Awareness
- 108 Chimneys & Flues
- 2K Domestic Hot Water
- 5.6K Gas Heating
- 105 Geothermal
- 160 Indoor-Air Quality
- 3.6K Oil Heating
- 69 Pipe Deterioration
- 956 Plumbing
- 6.3K Radiant Heating
- 385 Solar
- 15.3K Strictly Steam
- 3.4K Thermostats and Controls
- 54 Water Quality
- 43 Industry Classes
- 47 Job Opportunities
- 18 Recall Announcements