GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Integrating a Reed Bed System

eorl | Posted in General Questions on

Building a home and trying to build an integrated water cooling, reed bed system. Location has sub-tropic climate (Wet-hot summer/Dry-cool Winter).
Max temp = 30 Celsius (Dew point = 20 Celsius, Min Temp = 15 Celsius)

Basic design principles(Check Attachment) :
1. First tank is 8KL, made with concrete and underground with open top. This tank is 3.5′ deep and top 1′ is gravel/media material. Water level will be at 3′. So 2.5′ height of just water and .5′ height of water + gravel. This tank will have multiple aerators at the bottom to create movement and also to continuously improve water quality.
2. 2nd tank is 20KL. Same configuration as 1st tank in terms of depth, gravel, aerators and other parameters.
3. Water is continuously rotated between the two tanks plus the aerators around the tanks keeps the water disturbed.
4. Heat exchange will happen at 20Kl tank which is effectively 85′ long. Inlet and outlet is 6″ PVC pipe and exchange happens through 16X1″ PVC pipe laid in parallel near the floor of the tank.
5. Air is continuously recirculated from home through this PVC pipe system with outside air mixed at variable rate(depending on outside temp). Dehumidifier may be added inline to improve RH. Dew point does not look like a concern based on location weather profile. Air exits from the highest point in the home and cold air enters from under floor vents.
6. Home construction style is SCIP/3D EPS so I consider good insulation. Going for sealed windows and double glazed on South/West facing windows(North Hemisphere) to minimize load.
7. Area to cool is 2800 sqft with 10′ ceiling.

Technical parameters:
1. Tanks have 28 KL of total water in the system at any point of time. 3-4% of water will be refreshed every day. Most of the water change will happen during operation time. 1-1.5% of water will be lost by evapotranspiration(plants) and evaporation. Rest will overflow to percolator.
2. Total energy to raise 1 Celsius temperature of water is 32.56 KWh. Think I will have between 5-10 Celsius margin before I get concerned. So that is around 150-300 Kwh of energy that can be transferred everyday.
3. Ground heat transfer, evaporation and water rotation should keep the water from heating too much. Also think during night time with continuous water rotation, water temp should reset. Night lows are close to 15 and if I can take water even close to 20 every night, I should be good for next morning.
4. Evaporation of few 100 Liter of water everyday will not be an issue as can be compensated with daily intake of treated greywater.
5. PVC as a heat exchanger is not traditional but copper will eventually react with the not perfectly clean water(Kill plants). Also this research ( suggests that over time in recirculating systems the difference between PVC and metal is not substantial.

Decent livable environment. Not looking for 18 Celsius conditioned air but want the system to keep the home comfortable.

Would like to know the feasibility of the solution? The tanks will anyway exist as a way to have a tertiary water reservoir and maybe do some hydroponics on the top layer.

See attached image for high level design

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. Expert Member
    BILL WICHERS | | #1

    It sounds like you want to do space conditioning (air conditioning) of the indoor space with these tanks, so a sort of passive geothermal cooling setup? You are likely to get essentially no dehumidification with this setup, maybe a very small amount at best, which is going to be an issue with a high-humidity enviornment. You may be able to get some minimal cooling, but it depends on your average ground temperature, which I don't see listed here.

    The issue for cooling is that the efficiency of the cooling system decreases as the temperature delta (difference) between the "cold" medium that is doing the work of cooling and the air in the house that is being cooled. If you have a 20* delta, your system blows cold air and works well. If you have a 1 degree delta, you might not even notice that it's cooling at all. You can somewhat compensate for this by moving more air volume over the coil, trading "lots of slightly cool air" when you can't get "a small amount of very cold air", due to the limited thermal delta involved. Hopefully that makes sense. Without some type of heat pump type of system to force energy to move against the gradient (from a colder place to a warmer place, which is the opposite of what physics normally wants to do), you are limited in what you can achieve here.

    Note that it's worth mentioning that evaporative cooling is much less effective in areas with high humidity levels. Evaporative cooling works great in dry areas like Denver, but not much at all in humid areas like Houston. You mention you'll have "Wet-hot summer" seasons, which implies high humidity, and reduced cooling effectiveness.

    Aeration alone won't "continuously improve water quality". You probably won't get stagnant water with this setup, but you'd need more than just aeration. Putting some plants in is a good step, as the plants will help to strip nutrients from the water that would otherwise help less desireable things to grow.

    You could try a geothermal-type heatpump system, using the water tanks as a sort of evaporative cooler instead the usual buried cooling loop or wells. That would work, and would tend to keep the water slightly warmer than ambient, but would allow you more control over your indoor air temperature, and would allow for dumidification too. This is probably the route I would take in this case. I would not expect a passive system using just a heat exchanger, circulation pump, and fan to accomplish much here.

    Copper is pretty stable long term and probably won't be a problem here. Most water chemistries won't really react with copper on any high level. If that's a concern with particular water chemistry though, you can get stainless steel heat exchangers that will be more efficient than PVC, and won't react with the water. If you want the ultimate, you can also get heat exchangers that use titanium tubing -- which is essentially non reactive here -- but they are very expensive.


  2. eorl | | #2

    Hi Bill, really appreciate the thorough reply.

    Agree that reducing temp delta will reduce impact of the system. But delta between day and night is pretty high on hot days. High temp is around 30 Celsius and on these same days night low is 15 Celsius.

    Considering SCIP home construction and good insulation, I can use night to reduce temperature of air/structure inside home to as close to 16-19 Celsius as possible, even open windows and just go natural convection for the night. Along same lines water temp should also approach close to 16-19 Celsius by morning (maybe some open splashing fountain running in night to cool water). This should provide good thermal buffer/inertia gained in night both inside the home and within the water tanks to have a good enough performance during afternoon hot times.

    Along same lines when delta is really small, ambient temperature would already be close to 21-25 Celsius which although not perfect temp should be comfortable.

    Above points with the underlying assumption that I add a dehumidifier inline as that would probably bring the highest QOL improvement as function of power consumed in my location.

    As for suggestion to use a heat pump see my next post for underlying principles of the design. If I understand correctly even a self-designed heatpump system will rely on Freon to provide the needed heat transfer. That would be cost-prohibitive not for me personally but for overall principles of the design.

    1. Expert Member
      BILL WICHERS | | #8

      Sounds like this going to be more of a "thermal battery" sort of thing yet, which isn't what I had thought you were planning. I think you'll lose more water than you think though, especially with plants and transpiration. If you're worried about mosquito larvae, put some fish in the tanks to eat them. Fish make short work of those larvae.

      You will need to work out your time duration of night and day. You will probably find that the water tanks end up stabilizing at some sort of average temperature between day and night temperatures, and you'll see that temperature wander above and below that average slightly on a daily cycle. If the tank is small enough to cool off to the low temperature at night, then it is likely to also increase to the high temperature during the day, so you'll only have cooling capability for part of the day. You may be able to manage this by controlling the flow rate through your heat exchanger so that you maximize your cooling during the hottest part of the day, which is probably the simplest way to maximize your bang for the buck with this setup.

      Heat pumps do use refrigerant, not necassarily freon, which was only one type of refrigerant. I'm not sure why you're opposed to a system that uses refrigerant here though. A heat pump would allow you to cool continually throughout the day, warming the water tanks without stopping your ability to cool your structure. That seems like a plus to me, although it does require more energy than the passive system you're considering.

      BTW, black polyethylene sheet can have a long life too when not exposed to sunlight, and it's probably the cheapest membrane/liner material available, so that might be something to consider over EPDM.


  3. eorl | | #3

    Want to add a bit more on the philosophy behind the design. The reason to construct this design is to test cost/value viability in India/3rd world hot countries. Even though I am implementing this, my location is highly undesirable for this test as it has plenty of rainfall, good sanitation infrastructure and high humidity in summers. But large swathes of land in India/3rd world have water shortage and hot, dry summers. These same places also lack good sanitation infrastructure. During summer Swamp coolers act as an excellent low cost solution but the evaporation of potable water is a huge unseen cost. This loss can be as high as few 100 liters per day per household accounting for 60-70% of daily water usage for the household.
    With my home I want to design a system that recycles Greywater and "Actively" stores it for reuse. The tanks are basically post treatment horizontal recirculating reed beds capable of holding close to 2 1/2 month of greywater output(Just sink and bath, no kitchen so negligible bio load). This ensures that as a household all tertiary water needs(Gardening, car washing, flushing) are handled at net zero potable water cost. This is a very high level of water efficiency improvement for most places with uneven rain pattern across year.

    Aeration + continuous recirculation + abundance of plants(300-350 sqft usable area) + 2 1/2 month of HRT should keep water in good condition. On top of that since the tank is underground and has very soil like top, the design is pretty much invisible and mitigates any smell/mosquito larvae/insect related issues. If I face any smell issues, my first plan is to blend clean water and if it still persists, I plan to integrate ozone pump with the aerators(ozone over chlorine/UV as electricity is the only long term cost and ozone pumps are getting cheaper by the day).

    To me for a very large part of India/3rd world countries with water issues this is a net positive in term of QOL. Now I am thinking how to extend this design so that places that have dry hot summers and cannot afford to evaporate away 100's of liter of potable water can reintegrate this same design to provide further QOL improvement. The more layers of impact a single design can make, the more worthwhile the initial investment becomes which is a concern for people with limited cash. The design is highly synonymous with aquaponics/hydroponics for a reason as that can provide one more multiplying layer.

  4. Expert Member


    If you want someone knowledgeable who will talk you head off about this, contact Kim Rink.

    I went to school with him three decades ago. Great guy and absolutely obsessed by this type of stuff.

  5. eorl | | #5

    Hi Malcolm, Thanks for the direction. Emailed Kim and would love his viewpoint on different aspects of the design.

    1. Expert Member
      MALCOLM TAYLOR | | #6


      A couple of decades ago I helped install a wetlands septic system for a multi-unit residential development. It was a series of shallow ponds about 16" deep, lined with membranes, and filled with washed gravel. They were seeded with local marsh plants, and the effluent flowed through them 3" below the surface, which stayed dry. The marsh quickly thrived, became the home to all sorts of bird species, and the water that emerged was clean enough to be discharged into a nearby creek without any further treatment. I found the whole thing almost magical.

      I know that's only half of what you are attempting, and the other part may be quite challenging. If you manage to integrate the two it would be fantastic. Good luck.

      1. eorl | | #7

        I think membrane is a cheaper solution and EPDM can last as long as concrete but I am not onsite and want to go with local labor skill strength.

        Getting a lot of greenery, birds and very good quality tertiary water is win in itself for me. Just looking to enhance the value that can be provided considering 28KL of water represents a huge amount of energy reservoir.

        If it doesn't work I can reuse the infrastructure and install any inline HVAC solution but want to test the viability of the resource I already have for free. Using the system to pre-cool air before feeding it to mechanical system would produce at least some positive value. Total cost of installing the 16X1" PVC heat exchanger will be $200-300. Maybe even less so it is not a big investment but I can do some fun experiments and collect some good quality data.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |