Natural swimming pond in Thailand

[Meyer Jordan]

It is a mistake to operationally compare a 'regen' zone to a slow sand filter. The former utilizes up-flow which allows for a wide range of flow rates whereas the latter is downflow and relies on the hydraulic head created by the depth of water covering the top of the sand to determine flow rate.
Although slow sand filtration has proven effective in bacterial, viral and parasitic pathogen control in residential and small municipal applications, the sheer size of your application >1,000,000 gallons would make it very difficult to achieve the same level of disinfection in the swim basin. The slow sand filtration may very well perform as expected but only a very small percentage of the total volume of water will be treated at any given time only to then be re-mixed with the greater volume of untreated water. This effectively negates any expectation of achieving overall the disinfection rates attributed to a typical slow sand filter application.
FYI: The regen zones on NSP installations done by Bio-Nova and BioTop and subsequently adopted by most independent NSP installer are all upflow gravel beds.

 

[xye]

It is a mistake to operationally compare a 'regen' zone to a slow sand filter. The former utilizes up-flow which allows for a wide range of flow rates whereas the latter is downflow and relies on the hydraulic head created by the depth of water covering the top of the sand to determine flow rate.
Although slow sand filtration has proven effective in bacterial, viral and parasitic pathogen control in residential and small municipal applications, the sheer size of your application >1,000,000 gallons would make it very difficult to achieve the same level of disinfection in the swim basin. The slow sand filtration may very well perform as expected but only a very small percentage of the total volume of water will be treated at any given time only to then be re-mixed with the greater volume of untreated water. This effectively negates any expectation of achieving overall the disinfection rates attributed to a typical slow sand filter application.
FYI: The regen zones on NSP installations done by Bio-Nova and BioTop and subsequently adopted by most independent NSP installer are all upflow gravel beds.

It was interesting to read your mention that Bio-Nova and others mostly use upflow filters. I hadn't read that before.

When I first started designing our pond, I read several documents discussing both upflow and downflow designs and ended up going for a downflow design for a couple of different reasons. As such, I'm surprised that you categorically assume NSP's are nothing like slow sand filters because they are all upflow. I'm also confused why you say that NSP's allow for a wide range of flow rates when almost every document I have read that mentions the flow rates suggests that an NSP requires a controlled turnover rate of between 1-2 days, which is not a very wide range. Furthermore, I have done the math and found that this will always be on the low end of the rate that slow sand filters normally operate assuming one follows all of the design guidelines proposed by most NSP design books, Furthermore, an NSP regen zone besides maintaining a comparable flow rate will also inherently maintain the cover of water needed to keep the schmutzdecke alive. So I can't really agree with your assertion that an NSP's regen zone is incomparable to a slow sand filter, at least in the downflow case that I am using.

On the other hand, I completely agree with your second point. With an upper bound of .4m/hour of flow rate to maintain the schmutzdecke (biological filter layer) on a slow sand filter, a filter area equal to the swimming area (as is normally recommended), and an average water depth of about 2 meters, this results in a minimum turnover rate of about 5 hours. What this means is that only about 20% of the water in the pond will have been filtered in the last hour, at which point it has been mixed in with the rest of the water. Bacteria can grow quite a bit in 5 hours. This is my biggest safety concern in building this pond.

On the other hand, as I've mentioned above, for me it wasn't a choice between a swimming pool and a natural swimming pond. It was a choice between a giant stagnant pond, and an NSP design with a regeneration zone, active circulation, and slow sand filtration. So at least I am sure it is better than the alternative, whether or not it will be completely safe for swimming. And I will be getting the water professionally tested before I ever go swimming in it. (And probably continue to get it tested occasionally thereafter.)

I assume you are simply advising caution, which I agree is quite warranted.

BTW when I was reading up on safety concerns regarding NSP's I had read this article, which I thought was quite fair and balanced:
http://www.pondtrademag.com/pond-or-pool-the-fierce-and-divisive-swim-pond-debate/

I only just now drew the connection between the Meyer Jordan in the article and yourself. (I assume that it is the same Meyer Jordan as yourself.) Just for the record, I agree with many of the criticisms in the article about the professional risks of building a pond like this and then marketing it as "just as safe as a chlorine disinfected swimming pool". NSP's, by design, are not sterile. Chlorinated swimming pools are designed to be chemically disinfected.

It is just that my wife and I have never had any interest in having a chlorinated swimming pool on our land.

 

[Meyer Jordan]

I'm also confused why you say that NSP's allow for a wide range of flow rates when almost every document I have read that mentions the flow rates suggests that an NSP requires a controlled turnover rate of between 1-2 days, which is not a very wide range.

The first NSP constructed for public use opened in Minneapolis, Mn. yesterday July 24th. This is a 500,000 gallon facility constructed by BioNova. The declared turnover rate for this NSP is once every 12 hours or twice per day. This is a much higher rate than the 1 - 2 days turnover rate that you refer to.

What is the difference in elevation between your regen zone and the main pond basin? Hard to tell from the photos.

After looking at your photos again, I realized that the question of type of liner had not been asked. What type of liner are you using as I can see no evidence of liner in any of your photos?

 

[xye]

The first NSP constructed for public use opened in Minneapolis, Mn. yesterday July 24th. This is a 500,000 gallon facility constructed by BioNova. The declared turnover rate for this NSP is once every 12 hours or twice per day. This is a much higher rate than the 1 - 2 days turnover rate that you refer to.

What is the difference in elevation between your regen zone and the main pond basin? Hard to tell from the photos.

After looking at your photos again, I realized that the question of type of liner had not been asked. What type of liner are you using as I can see no evidence of liner in any of your photos?

Actually, based on a similar rough calculation like I did above, a 12 hour turnover rate would result in about a .17m/hour flow rate, which is quite consistent with a slow sand filter. Normal flow rates are in the .1-.4m/hour for slow sand filters, but I've seen studies that suggest slower is better. (I've seen numbers down to about .04m/hour.) So it sounds like even an unusually high flow rate NSP still stays within the slow sand filter range.

This is a helpful data point for me. Thank-you.

I am really hoping to avoid getting into a debate on the technical merits of NSP's here. I'm just trying to get a bit of advice about how to make the one I'm building as successful as possible. But you clearly know a lot about NSP's (and probably ponds in general) and I do appreciate the information you are sharing. Thank-you for sharing it. And if it helps to put your mind at ease, please imagine I will never ever swim in it. As I mentioned, this pond is primarily a giant garden pond for landscaping purposes, but I would prefer to circulate and filter the water rather than just leave it stagnant. NSP's provide a pretty good knowledge base about how to do that in a way that won't kill off all of my plants. I'll decide about whether to take an occasional dip in it after it is done (and after I've done rigorous safety testing.)

Moving on, I think that I have a pretty good sense now about the water circulation flow rates, but I still have questions about the aeration air pump flow rates that I'm hoping someone might be able to help me with. (Even example numbers that other people are using would be helpful.)

For context, my current best estimate of the pond volume is around 2500-3000 cubic meters (660,000-800,000 gallons). I'm uncertain of the exact volume of water that the pond will have since there is some uncertainty about the displacement of the filter sand and the sand will be contoured for various water depth plants.
(incidentally, this volume is quite close to the 500,000 gallon facility Meyer Jordan mentions above, so any available numbers about air pumps from that pond could be a pretty good starting data point here for me. I will try googling it!)

I remember reading in one of the NSP design guides that a 250W air pump should be enough for larger ponds up to about 1 acre. (Mine is about .45 acres.) But I think this is a bit of a silly statement. Air pumps are rated by their air flow rate at different reverse pressure amounts. Based on my understanding of centrifugal pumps in general, I would assume that a 250W air pump designed for high flow rates in shallow water might not work at all in deep water while a 250W air pump designed for deep ponds might have less flow rate in shallow water but still work in deeper water. In my case, I am hoping to put the air stone (stones?) at about a 2m (6.5ft) depth. This comes out to about 2.9 psi backward pressure for the air pump.

Can anyone recommend how much aeration air flow rate I should try to achieve? What kind of air pump would I need to achieve that air flow at 2m depth? (I actually bought a 250W air pump already for testing... I'll try to post some pictures of it later.) How many separate air stones do you think I should use?

What do most people use for aerating their garden ponds? (Please include the size of the pond, if possible.)

 

[Meyer Jordan]

Here is what is supposed to be a rough schematic of the Webber Park NSP in Minneapolis. I do have a more detailed copy somewhere on my computer if I can only find it.

You will notice that the aeration is pumped directly into what they call the filtration shaft and is apparently not a standard bottom diffuser or airstone configuration.

Just curious. Considering you stated goal(s) of this project, it seems that it would have been somewhat less expensive to go with the upflow regen zone (which is basically no more than a sub-surface flow constructed wetlands) rather than the slow sand filtration which will also require more maintenance.

 

[Meyer Jordan]

I found it!!
Here is a more detailed schematic of the Minneapolis NSP. I think that you will find this interesting.

 

[xye]

That schematic is quite interesting, and I guess one of the most interesting thing about it is how complicated it appears to be. I remember from one of the articles I read recently about this particular pond (after you mentioned it) they talked about an upflow biological filter shaft. Now that I see the schematic, I think I can visualize it much better.

I'm definitely trying for a much simpler design. If I can, I'll try to draw up a presentable schematic of what I am working towards over the next few days to open it up to any suggestions. But the basic idea was to target simplicity of maintenance over reducing the amount of maintenance. While I don't have much experience with this particular kind of project, the various engineering projects I've done over the years (soccer playing robots, autonomous vehicles, computer vision systems, security systems, electric vehicles) has taught me that when you are doing something for the first few iterations, many things will go wrong. The designs that end up working best in the early iterations aren't the most sophisticated, but rather the ones that make it easiest to fix the things that go wrong. I guess if you are BioNova, you've gone through these early iterations and worked many of the engineering kinks out and can implement more advanced designs.

In the beginning I was actually planning an upflow system just because that was what the first DIY resources I found seemed to show. Then I ran into a few downflow designs and realized I liked them more. This page has a simplified schematic I stumbled across a while back:
http://www.ecohome.net/guide/natural-ponds-natural-swimming-pools

For the pond design, I reasoned (rightly or wrongly?) that since on an upflow, you are pushing water into the filter from small holes in a pipe at the bottom, you would have a greater chance of particles clogging up the pipe. And the clogs would be in a very hard to reach location. Thus the backflushing. But I always worried the backflush would only unclog a small percentage of the holes, leaving it susceptible to clogging again in short order. In a downflow, the particles "enter" the filter on the surface, and thus should get trapped somewhere in the much larger area of the top of the substrate, which is unlikely to ever clog. And any buildup on the top of the substrate will be easily accessible for maintenance and cleaning anyway. Put another way, only filtered water ever touches the hard to reach parts of the pipe.

As a secondary advantage, the general flow of water on the surface would be from the swimming area to the planted area, so this might have a slight "retaining" effect on the floating plants, keeping them in the planted zone. I'm not sure if the flow rates are enough, though. A rough calculation suggests an average of about 1.5cm/second of water flow rate above the retaining wall based on my pond configuration if I target a 1 day turnover rate.

I suspect that the design I went with is really the same as the subsurface flow wetlands you described, just with the water direction reversed. I have perforated pipes in a gravel layer meant to allow the water to flow horizontally toward the pipes after the water seeps downward through the sand. If the pump were reversed, the gravel layer would allow the water to flow horizontally away from the pipes to spread out before seeping upwards through the sand. Or am I missing some more fundamental conceptual difference here?

As a side note, the "Thai long tail boat propeller" based circulation pump prototype we've been building is almost ready for testing. The entire thing has cost in the range of $50-$100 so far, including custom machining for couplings, etc. If I'm lucky, I'll be able to test it by the end of the week. I'll try to post some pictures. (And numbers.) Because it is only circulating the water horizontally rather than pumping upwards, I'm estimating the the large 12" propeller will only have to rotate at around 100-150 rpm to get the approximately 30 liters per second (30,000 gallons per hour) flow rate I'm targeting. The large, slow impeller should result in much higher efficiency. And maybe it is less likely to get clogged by any small debris that does happen to end up in the pipe. (Small pieces of gravel that make it through the holes?) At least that is the theory. We'll have to see for real when it is operational.

Also, does anyone know where to source a cheap "river water flow meter" that I can use to test the flow rate at the exit of the circulation pump? The only ones I found for sale here in Thailand were around $7,000 when I asked for a quote. It looked like a toy boat propeller on the end of a stick connected to a simple magnetic encoder. Here is a link to a similar one:
http://www.vernier.com/products/sensors/flo-bta/

I suspect we could build a simple one for around $20-$30, but this will take some more time. (And will need to be calibrated somehow.) I really don't need a $7000 precision instrument. Is there any cheap tool like this for sale on eBay or anything? I looked and couldn't find one. The only ones I saw are the kind that go in a small pipe as part of a household water system.
http://www.ebay.com/bhp/water-flow-meter

II can't think of how these can be made to work for measuring flow rates in the pond. :-(

 

[xye]

I found it!!
Here is a more detailed schematic of the Minneapolis NSP. I think that you will find this interesting.

I somehow missed this more detailed schematic earlier. It is quite interesting, indeed. Thank-you so much for sharing!

The surface area of their swimming zones, when put together, is roughly double the size of mine. But I think the average depth is substantially less. I added up their deep and shallow water pump flow rates (and converted into metric... sorry) and found it comes out to around 60 liters per second. This is only slightly higher than what I will be trying to get (total) from my main 3 circulation pumps.

I feel like I'm a bit too much of a beginner in this to pass judgement, but I must admit, there is a bit too much "sleight of hand" in their design for my taste. There are flows of water going every which way. For example, why does the "water fall" pump take its water from the "dirty" water tank rather than the "clean" water coming out of the "regen" zone? (Like all the other swimming zone water inlets do.) And the waterfall looks like it is designed to appear like water from the upper shallow pond is pouring into the lower pond, but it is actually being pumped from somewhere else completely. Oh well, maybe best not to dwell too much on this. I'm sure there are design goals/constraints I'm not aware of. And they had to spend that $6.5 million dollar budget somehow, right? ;-) (I think I read that was the number somewhere.)

On the subject of budget... how much do people normally spend on these kinds of ponds? I think I've cut some expense with the DIY approach (and maybe cut some corners on pond liner, etc), and given the cost of labour in Thailand, I think I'm doing this quite on the cheap. (Yet another reason to be cautious about treating it as a swimming pond rather than just a decorative pond!)

Which brings me back to the topic of liners. As you observed from some of my pictures, I'm using a natural clay lining. Based on my early price estimates, this cut the entire price of the project in half and reduced the complexity substantially. The soil here has a ridiculously high clay content, and once I dug down below the topsoil, it had the consistency of, well, wet potting clay. (No plant matter at all.) Rain water puddles don't sink into the soil at all over observation periods of weeks. And for the first week or so until the algae begins to grow, the rainwater puddles are almost perfectly clear. Even intentionally disturbing the water and trying to kick up some turbidity by stirring up some of the clay, it settles back to being clear within a few hours. And even when I dug down 2-3 meters deeper than the surface of the canal just 10 meters away (at its closest point), there was no observable inward groundwater seepage for the month or two before the rainy season came (at which point I wouldn't be able to tell any more).

Are there any problems associated with a natural clay lined pond other than possibly increasing the turbidity? Does anyone think this choice was a mistake?

 

[Meyer Jordan]

The waterfall in this scenario requires a dedicated pump because a higher flow rate is required to create the visual impact desired.
I suspected from the photos that this was an earthen-bottom project.The potential problem that I see here is an eventual build-up of silt in the regen zone.
If the overall flow pattern from the main pond basin is over the retaining wall separating this from the regen zone, What will be the force behind this flow? Will you be drawing water from the base of the regen zone through the use of pumps? If so, you may want to rethink this configuration.

 

[xye]

In case anyone is interested...

We've gone ahead and built a really easy and cheap water flow meter. We spent about $4 building it, but we had tools and parts laying around.
I tried attaching a few pictures of it, but the "Upload file" process kept giving me errors. So here is a link to my blog post on the subject:
http://mibproj.blogspot.com/2015/08/ultra-cheap-water-flow-sensor.html

Now I have a way of measuring water flow for the DIY circulation pump we are building.

And the DIY pump has been tested and appears to have been able to circulate about 45 L/sec using only 36W of power. And using a DC motor, it should be fairly easy to hook it up to solar panels later. Here is a link to my blog post on the axial flow circulation pump:
http://mibproj.blogspot.com/2015/08/first-in-water-test-of-axial-flow-water.html

(Lots of pictures on those pages.)

Update: After posting the reply, I then went to edit and was able to upload photos. See below. Not sure why it was different. Oh well, no big deal.

 

[Chelseabreeze]

Hi xyi. I just completed the construction of my organic swimming pond in Chelsea, Quebec, Canada. After 3 years of research of various methods, I opted for the design concept by David Pagan Butler in the UK. David employs the use of a low pressure air compressor driven air stone 'bubble lift pump' system to move water. Pond life...plants and critters...appreciate low pressure. I followed his process and it is working out remarkably well and is very cost-effective and low maintenance. I encourage you to research his process. He offers a comprehensive guide and DVD. I just posted my experience complete with some photos here...
https://www.gardenpondforum.com/threads/my-organic-garden-swimming-pond.17687/

I wish you good luck, and no injuries through your process. I survived unscathed...barely a scratch.

 

[xye]

After 3 years of research of various methods, I opted for the design concept by David Pagan Butler in the UK. David employs the use of a low pressure air compressor driven air stone 'bubble lift pump' system to move water. Pond life...plants and critters...appreciate low pressure.

It is interesting you mention this, since I just stumbled across his website about a month ago when I was looking for answers to some of my questions that came up in the middle of construction. I immediately bought the PDF book and I must say that it is the most forthcoming of the books on these kinds of projects that I've found. Most resources I've read seem to focus on the aesthetic assuming you are not going to do the engineering yourself. But for water flow diagrams, wiring diagrams and safety principles, the bubble pump concept you mention, and DIY skimmers and overflow pipes, that was the best book on the subject I've read.

My pond design actually has two "water zones" on opposite sides of a "plant zone". The smaller one is more of a calm "water garden" area and probably has about 25cu m of water. For that area I decided to adapt my design to try David's bubble pump idea pretty much as he described it.

But for the main "water zone" I'm developing a water circulation pump based on readily available long tail boat parts here in Thailand. This zone probably has about 1500cubic meters of water so it needs a lot more circulation. In a recent test I did I've found that I can get about 45L/sec (~11gal/sec) of flow from only 36W of electricity. This is about 100x the efficiency (for very low heads like a circulation pump needs) compared with the 200W aquarium pump I tested before. Since my goal is to power the pumps using solar, 36W vs 3600W is a very very big difference. (Per pump and I need 3 such pumps.)

And I can keep the electrical parts out of the water and drive the pump with a rubber belt, which should have some of the electrical isolation advantages of the bubble pump idea. Also, the low pressure concept of the bubble pump that David pushes is exactly what I was going for in my "long tail boat" based circulation pump. Because of the very large size of the 12" impeller (the long tail boat propeller I bought) it only needs to rotate at about 2 revolutions per second to achieve these large flow rates, so even small-to-medium sized leaves can pass through without getting caught.

Of course, my pump prototype does require welding, machining, and more electronics resources not to mention the ready supply of cheap long tail boat parts here in Thailand (I spent about $30 on the parts) so it isn't really something I'm suggesting for everyone. That is one reason the "bubble pump" is so great in terms of the simplicity of design.

Just out of curiosity, do you have any way of measuring the water flow rate of the bubble pump? What about the power consumption?