Monday, January 6, 2014

Bed Rock: Design

Efficiency is logic, but life is experimentation.  In the Spring, I rented a bedroom with no heat or electricity, with the intention of retrofitting before winter.  For well over a decade I have rented spaces that are antiquated, where winters are uncomfortable, drafty, expensive.  Renting is great, convenient, except when you want to fix the problems you face with an old house.  The following post describes my design process for my curious heated concrete bed project.

BACKGROUND:
In a heat transfer course, circa 2004, Dr. Ray Francis used some math to demonstrate how much more effective and energy efficient radiant heating is compared with forced air heating systems.  Despite my intense interest in hydronic heating, I have been distanced from such systems due to socioeconomic realities.  Hydronic heating installations are often motivated by comfort, and its owners are often faced with more pressing matters than fussing over energy savings-so we don't hear about it.  Often the capital cost of hydronic heating yields long payback periods on energy savings, and prevents installation by lay persons.  The parts involved with heating and circulating liquid phase water seem relatively simple, and I suspect the lacking demand within the hydronic marketplace is the major expense of such installations.  My unique living situation has afforded me a golden opportunity to create an experimental hydronic heating system, and to learn by doing.

THE RADIATOR, THE HEAT SINK, THE INTERFACE
While I mixed concrete, Pa helped
with his garage and trowel
In my research, I discovered a product called warmboard, which I found exciting for the growing popularity of hydronic heating retrofits and new installs.  There was a lot of temptation to utilize a single 4'x8' warmboard as my radiator.  Warmboard countered my inital design concept of a "thermal mass" of rock.  Warmboard's website informed me that a hydronic system with a continuous heat supply operates more efficiently with a conductive (thin metal surface) radiator than with the more classical thermal mass (big warm rock) radiator.  While exploring such concepts I gathered that an intermittent heat supply, as in solar heating, is best coupled with a "thermal mass" radiator.  My first installation will employ a constant heat source, but will only cover 1/3 of the rooms total floor space.  Because my installation is not permanent, and heat source flexibility has exciting future applications (van heater, garden greenhouse/sauna, next dwelling, etc), I opted for a thermal mass radiator design.  The slab was poured in three sections for mobility.  The mass came in around 600 lb.  The top surface is close to 25 SQ FT.  Unlike typical flooring slab radiators, my slab is suspended 1 inch from the floor.  The initial purpose of suspending was for leveling of the three distinct slabs.   I intend to sleep on this surface, which will limit the slab's room heating capacity by being insulated with bedding.  A serendipitous benefit of suspending the slab may be additional room heating via convection from the bottom surface.

This component needed to be custom crafted, and the construction method was adopted from the growing popularity of concrete countertops.  The hardware store stocks a 90 lb premix concrete bag for this purpose, Maximizer 5500+.  This mix is a high bulk (low density) concrete, which is less ideal for thermal conductivity, but helps reduce the load on the building structure, vehicle, and handlers of this material.  The final thickness came in at 2.25", to which I would design another .25 or .5 inch next time for wire mesh wiggle room (wire mesh comes from a roll, and isn't flat).  I laid the mold on a sheet of plastic laminated, 1/8" masonite, to obtain a smooth surface finish.  In addition to smoothness, I was pleasantly surprised with some curious bubble formations, due to water creeping into the laminate.

FUEL
One beautiful aspect of hydronic heating is that a wide variety of fuel systems can be applied by simply re-plumbing the system (fuel oil, gas, electricity, solar, geothermal, automobile coolant, etc.).  Someday, I want to utilize solar energy to heat my residence via hydronics, but for now, the convenient, experimental fuel delivery method will be coal fired electricity.  I did look into solar energy availability on SLC's solar map, and didn't find much potential in the immediate region of my heating needs.  The big decision here was tankless vs tank heater.  Conceptually, a thermal mass (big warm rock) hydronic system serves as a storage tank for heat, and the tank-ed water heater is somewhat redundant and highly space consuming.  Space savings, and the option to use energy saving 220V power supply pushed this decision in the direction of a tankless heater.  I imagine a hybrid solar-house powered hydronic heating system could certainly benefit from utilizing a tank-ed hot water heater for the built in energy storage capacity.


PUMPING
To transfer the heat within a hydronic heating system, water circulation is critical.  The use of water as an energy carrier conserves a great deal of space and energy as compared with forced air heating systems.  For space, compare cross sections of 1/2" water hose to 1ft square aluminum air ducting.  For energy, pumping water slowly (2GPM) through a hose is dwarfed by the energy demand of blowers conveying hot air rapidly through heating ducts.  Because my slab radiator is smaller than most, I carefully selected a hydronic circulation pump.  The selection was Tyco's smallest (006) closed loop hydronic pump for the interest of proper flow rate and noise minimization.  Taco pumps have a reputation for being remarkably quiet, but an improperly designed conveyance system can produce noise via fluid flow.  An incredibly helpful resource in this project has been http://www.climatemaster.com/downloads/RP878.pdf.  Utilizing this source, and Taco's published pump curves, I was able to construct a graph, and select a design that employs an appropriate fluid velocity to minimize air bubble formation, air bubble growth, fluid noise, all while optimizing heat transfer rates.

PLUMBING
PEX (cross "X" linked Poly Ethylene) seems to be the unanimous choice of transport media for most plumbing needs, but there is a specific PEX designed for hydronic heating-Oxygen barrier PEX.  Apparently oxygen can penetrate standard pex hose, and this can cause problems in a closed loop heating system (bubbles, rust, grit, concrete degradation etc).  The smallest oxygen barrier PEX tubing available is of a 1/2" diameter, but all my other plumbing fittings (pump, air eliminator, pressure relief, etc.) use 3/4" fittings.  An air eliminator was installed to keep the air bubbles and oxygen and rust to a minimum.  I anticipate draining, flushing, relocating and/or modifying this experimental system frequently, so threaded steel gas pipe was used for plumbing connections.  The equilibrium flow rate is anticipated to be on the higher end of ideal (2-4 GPM) for my 1/2" pex system, so I've added a globe valve after the pump to help increase the resistance, and limit flow rate.

METERING
Much of this project's design parameters came from http://www.climatemaster.com/downloads/RP878.pdf.  To validate my parameters, and monitor my system controls, I'll need to make some measurements-flow rate and heat input should help.  Two analog temperature gauges will monitor intake and discharge temperatures for a heat balance from the heat source.  Flow measurement will be gathered indirectly from pump curves, using two analog pressure gauge measurements before and after the pump.   

2 comments:

  1. Evan, thank you for the kind comments on that 'other' site. You will be glad to know we now have used the cheaper tankless water heaters (in lieu of the more expensive boilers) for over three years now with no problems - and have not had to use our backup heater - it is still in the box. No 'legionnaires disease', none of the problems the 'professionals' on that site warned about. Maybe we have just been lucky, but I got the feeling there was not enough profit in this system for the 'pros' to promote or back the concept. After all the bashing, I saw no reason to continue posting on that site. Taco does make a smaller pump than the 006 and with no rise you could go with a minimal unit in your situation. Looking at your photo above, I think I would have not skimped on Pex and would have doubled up (every three inches), but it should work fine - it is just as I laid it in the floor of our residence. If you use a heater that accepts hot water it should serve you well. Dave Eveland.

    ReplyDelete
    Replies
    1. Dave, no sweat. I appreciate you sharing your experim-ence with such systems-your testimony was pivotal in keeping my morale fueled to see this project through. After burning up 2-3 Eemax tankless electric heaters, I ended up having success with a Rheem RTE-7, though getting it started required adjusting the flow switch to a more sensitive location-accomplished simply with a set screw. 7 KW is a bit more than my energy needs, but with the added current control of the thermostat, the added capacity might not dictate excessive consumption in this case. I toyed with on/off control, but got a brief steamy/percolating noise in the heater upon startup. I'm a bit shy of this noise after burning several previous heaters. I have found reasonable control by simply setting the thermostat built into my Rheem heater. I am finding simple comfort by keeping the circulating water temp in the 90F range. Winter wanted 4-5 of the Rheem thermosat tick marks, presently it wants only 1-2 tick marks. The Taco circ. pump runs all the time, and it feels hot, but it keeps going, and draws little electrical current. Yes, my slab could benefit from closer pipe spacing, but honestly, the radius I used to bend the 1/2" Oxygen barrier pex seemed close to kinking. Maybe pre-heating the tubing would make for tighter radius bends. I don't anticipate running this heat system trough the hot, dog days of SLC summer, but it surely seems convenient to plug in a pump and have instant heating capability.

      Delete