Rewind

It is April 27, 2010 and it snowed all morning. We've also had several days this month with temperatures in the 70s -- it even touched 80 degrees one day, the pond reached 65 degrees, and the kids went swimming. But today it is snowing, in a demonstration of why April is the cruelest month.

In commemoration of the last snow of Spring and anticipation of the warmth to come, here are a couple of photos of Winter pond activities. These represent an almost daily after-school occurrence -- cleaning off a light covering of snow, followed by skating or sledding. Thanks to an early stretch of cold weather, the ice was thick enough to support us well before Christmas.

It's only water, we can out think it!

Once the pump was plumbed and all the cemented joints had had time to set, I was eager to finally see water flowing.  I filled the pump's filter basket with water to prime the pump, and plugged-in the pump in via an extension cord from the house.  The pump kicked on, whirring away quietly ... but no water flowed, even after letting the pump run for a couple of minutes.  Although the pump was advertised as "self-priming", it was unable to pull water the 2' to 3' vertical distance from the existing water level.  Dumping several buckets of water into the pump had no effect -- the water just flowed down the suction lines until it reached the pond's water level since there were no check valves installed in the suction lines.

It seemed pretty clear that I needed to install check valves in each of the suction lines below final water level, which were all under several tons of rock.  Fortunately, we hadn't finished filling the pond yet, so I could still install the check valves at a point above the current water level that will be below the final water level.  Otherwise, I would have had to drain down the pond.  

After an hour of excavation, I had located all five pipes. I was pretty sure that the pipe on the far left was a discharge line, but had to cut it anyhow to make sure -- it wouldn't be good to put a one-way valve pointed the wrong way in a discharge line.  A lungful of air and bubble emerging from the deep zone confirmed that this was a discharge line, so I put this line back together with a coupling.  The remaining four lines had to be on the suction side, so I put a check valve on each.  This process actually took over two weeks since the supplier initially sent the wrong valves, and then sent the correct valves in a decidedly lethargic manner.

After that, priming the pump was much more effective.  The check valves prevented priming water from draining down and, after turning on the pump, water began flowing almost immediately.  Once I determined everything was working correctly, I reburied all the exposed pipes.  At this point, the pond was not filled to its final level so I called the water guy.

A few days later, and the pond is close enough to full that I can open the valve for the skimmer suction line.  This, however, revealed the second problem, which was that water would not draw through the skimmer.  Examination of the weak points of this circuit showed that all connections were tight, and I began to wonder whether I had installed a check valve backwards.  Before excavating all that rock again, I put a garden hose into the suction pipe in the skimmer and turned on the water in hopes that the line was air bound.  After a while, the skimmer did begin drawing water, but extremely weakly.  In fact, the suction was not enough to draw anything into the skimmer, which sort of defeats the whole "skim" part.  The only explanation I have now is that, because I have four lines going into one manifold, water is being drawn through the path of least resistance, namely the three submerged regeneration zone lines.  To correct this I will have to reconfigure the suction lines so that the suction between the regeneration zone and skimmer is balanced.  I'll probably feed all three regeneration zone lines into one 2" pipe which will then be connected to a manifold.  Further, both the skimmer and regeneration zone lines will be valved so that one can be throttled down to increase flow through the other.

So, that's pretty much the way not to do things.  Here are some recommendations on what to do:

1. Check valves at or below the final water level in suction lines.
2. Connect together all suction lines from the regeneration zone below the water line somewhere in the pond.  This will provide for better balance between the skimmer and regeneration zone, and allow you to use just one check valve for the regeneration zone.  Also consider adding a valve where the regeneration zone line connects to the pump in case that line needs to be throttled down to improve suction through the skimmer.
   
3. In cold climates, use threaded unions to make connections to the pump to make it easier to remove all water from lines and the pump above pond water level. 

Pond Pump

Now that we have the regeneration zone, skimmer and waterfall set and plumbed, it's time install the pump. I chose a Sequence Primer 7800PRM24 high efficiency pump. The goal for natural water filtering is to turn the entire pond volume over four times each 24 hours.  If you know (or calculate) your pond volume, multiply that number by four and then divide by 24, you will have the gallons per hour ("gph") you need to pump. Pumps are rated by gph at their maximum flow, but "head" must be taken into account to determine what pump you need. The more pipe, bends, valves and elevation the water needs to pushed through, the lower the pump's actual gph capacity will be. So even though our pond requires about 3700 gph to be moved, a much larger 7800 gph pump was called for.



First item of business was to pour a pad on which to set the pump. We used 2x4s to build a form directly behind the waterfalls, and mixed concrete in a wheelbarrow -- piece-o-cake! In a rare moment of forethought, I had previously labeled the tangle of pipes that need to be connected to the pump before they were buried in gravel (see photo below). Although the flexible PVC pipe is easy to layout (especially if you let it warm in the sun first), it is difficult to join it to PVC (Schedule 40) fittings. I found that using purple primer, as you normally would, prevented the PVC glue from reacting with the flexible pipe -- fittings could be pulled apart from the flexible pipe even after being allowed to cure for hours. After experimenting with different types of PVC glue, the best joints were achieved with regular medium duty PVC glue and no primer.

Before making the connections to the pump, I built a frame for the pump house (and I should have sheathed it with plywood -- would have saved some time cutting around all those pipes later on). Using several Schedule 40 3" x 2" T-fittings, I made a couple of manifolds to accept four 2" pipes on the suction side (3 from the regeneration zone and one from the skimmer) and two 2" pipes on the discharge side (for the waterfall and "stirring" line in the deep zone).  I put a ball valve on the skimmer line for when I want to take that off-line, like when the pond surfaces freezes over. Likewise, I put a valve on the discharge line to the deep zone in case I want to throttle that down and thereby increase the flow to the waterfalls.  More on stuff I should have done in "It's only water ..."

Snow Day

We awoke Friday (October 16) to three inches of snow ... the first of the season and about four weeks early by my reckoning.

Waterfall

In a swim pond or koi pond, waterfalls are not just for looks or the cool sound of water splashing on rock.  They also filter and aerate the water helping to keep the pond water sparkling clean.  In fact, they are often referred to as "filter falls" or  "biofalls."  I bought an Atlantic BF2600, which is basically a very sturdy plastic tub with a 26" spillway.  It comes with a grate that fits inside several inches above the bottom, two filter pads that rest on the grate, two "media" bags that hold the filter media of your choice and rest on top filter pads, and another grate for the top on which you place rocks to help disguise the waterfall tub.  

Similar to the gravel in the regeneration zone, the filter media will be a home to beneficial bacteria that will convert dissolved nutrients.  I read that lava rock was a good media. Luckily, it is available at Lowes as propane grill briquettes over in the outdoor living department.  Four 7 pound bags seems to make a good layer in my falls.

Water enters the waterfalls through a 2" bulkhead fitting (also included) at the bottom.  As with the skimmer, you drill the hole for the bulkhead fitting depending on which side of the tub is convenient for your set-up.  I put it on the back since the pump will be behind the waterfalls.  Also like the skimmer, the pond liner is attached to the face of the waterfalls to keep the water in the pond where it belongs.  Again, the manufacturer supplies a faceplate, silicone and all the hardware needed.  If the waterfall is to be placed outside of the pond liner, additional liner is used to contain the water in a stream that flows to the pond.  Unlike the skimmer, however, you don't have to use a transit to calculate water levels --  just place the waterfalls on level, firm ground wherever you like.  The only calculation needed is is to figure the head -- if the falls are too high, the head may be so great that your pump may be pumping far less water than what you were counting on.  Most of the pump manufactures' web sites will have information on calculating head and pump outputs at various heads.   

Once the falls are in place, it's time to create.  We started with a heavy layer of the cobble, and then placed a couple of large, flat rocks donated by Brian donated. The one on top was really heavy and had to be placed with some chains and the backhoe.

Water is either stupid or obstinate, and won't necessarily go where you want it to. About half the volume of water falling on to each of the large rocks was running down the backside of the rocks and disappearing into the underlying cobble, greatly decreasing the whole visual effect of a cascading water feature.  To help keep the water moving in the right direction, Atlantic makes an expanding foam which is just like "Great Stuff" except that it is black, making it practically invisible when water is flowing over it.  I used it to set smaller rocks and fill in voids between rocks, and that got all the water spilling down the face of the waterfall.  The photo below shows our initial effort before using the foam. Additionally, the water flowing out of the filterfalls is about half the output of the pump -- the other half is being directed through a 2" pipe into the deep zone so that water does not stagnate there. The photo also shows the pump house and the barrow that will conceal it -- more on that stuff later.

Skimmer

While the regeneration zone will filter small particulate matter, it isn't very good at removing things like leaves and grass clippings that float on the surface of the water which will eventually sink and litter the bottom of the pond, decompose and provide nutrients for algae to grow on.  That's were the skimmer should help.

Based upon a SWAG, I purchased an Atlantic "Big Bahama" PS7000 pond skimmer.  It is rated for ponds of 900 square feet of surface area and flows of 4000-7000 gallons per hour.  During the warmer months of the year our prevailing wind is from the south-west, so I decided to position the skimmer on north-east corner of the pond in hopes that surface debris would be propelled in the direction of the skimmer.  We pulled the liner away from that corner of the pond and Brian excavated a hole for the skimmer.  He used a laser transit to set the depth so that final water level (which is indicated by a raised line on the face of the skimmer) will be a few inches below the crest of the pond liner, which is about 18" above the top of the deep zone walls.  

Before placing the skimmer, I drilled holes on the skimmer walls for the 2" suction pipe, a 2" overflow pipe and an Atlantic Auto-fill valve.  The drill points had already been marked by the manufacturer so no measuring was necessary.  The auto-fill valve is similar to a toilet fill valve -- basically a float that allows water to pumped in to the skimmer when the water level drops, and shuts off water when the proper level is reached.  We ran some 3/4" water line in a trench from a nearby spigot on the house to the area of the skimmer, and used garden hose on each end to make the connections to the spigot and the skimmer.  

The overflow pipe on the other hand, which is about 2" above final water level, will allow excess water to gravity drain out of and away from the pond during the spring thaw or after a long period of rain.  It consists of just a length of 2" flexible PVC which we terminated with some rigid 4" pipe, which you can see here where it daylights outside of the pond.

The suction pipe is connected to flexible PVC and runs under the regeneration zone to the pump.  More on that when we get to plumbing the pond pump.

After this prep-work, the skimmer could be placed in its hole and the liner draped back over it.  A hole was cut in the liner for the skimmer opening and then the liner was attached to the face of the skimmer with a face plate, screws and silicone which were supplied with the skimmer.  We then backfilled the hole with soil and gravel.  This photo show the guts of the skimmer -- the suction pipe (which extends towards the bottom of the skimmer), clockwise from that is the bulkhead fitting for the overflow, then the autofill.  You can also see the skimmer's net just inside the door and its five brushes, which filter large debris from the water before it passes into the suction pipe.  

Regeneration zone

Up to this point, I have been presenting things in chronological order.  Now that various bits and pieces are being worked on simultaneously, it'll be easier if I just talk about one aspect at a time.  Let's start with the all important regeneration zone.

The regeneration zone surrounds the swim area and will handle the bulk of the water filtration.  This occurs as pond water is drawn through a layer of gravel, which eventually will be planted with a number of aquatic plants.  The gravel traps things in the water that don't float (like soil) filtering them out of the water.  It also provides a home for beneficial bacteria that will consume much of the nutrients in the water so that algae will not have as much to live on.  The aquatic plants will also take up nutrients and provide some shade, further depriving algae of nutrients and sunlight.  That's the theory, anyways.

The first step was to finish lining the pond by splicing the extra 10' width of liner.   Aquascape makes EPDM splicing tapes -- a 3" double-sided tape that adheres the two sections of liner together, and a 6" single-sided tape to cover the resulting seam.  The tape is made of the same material as the liner.  Although the instructions tell you to work on a flat, clean surface, that wasn't possible here.  Working slowly over our lumpy, creased liner, it took two of us about 2 hours to join the two liners, and then it took me another two hours to apply the 6" cover tape.  Note that the instructions "recommend" using a primer (also available from Aquascape) on the portions of the liner to be spliced/taped.  I would say the primer is mandatory for a good seam.  It makes the liner tacky and bonds the tape so well that it is impossible to pull apart.

Next, we put a layer of filter fabric over the liner in the regeneration zone to protect it from the 60 tons of gravel that will be dumped on it.  As you can see, our last load of water was enough to put several inches of water in the regeneration zone, which helps press the liner into the outside of the deep zone walls.

Finally, it was time to do some plumbing.  In order to draw all that water through the gravel, three 50' lengths of 4" drain tile were placed in the regeneration zone.  I used a variety of fittings to connect the drain tile to three lengths of 2" flexible PVC pipe, which will eventually be connected to the suction side of the pond pump.  The drain tile floats, so some rocks were placed on it to weigh it down.

Time for the gravel.  The bottom layer is 2" to 4" natural cobble, meaning that it is a smooth river rock as opposed to the jagged crushed rock.  Brian skillfully dumped gravel into the regeneration zone with his loader as I sat on the deep zone walls simultaneously holding the drain tile down with my feet and pulling up on the filter fabric so that it didn't get pulled away from the liner.  After each two or three loads, we would push the rock by hand into an even layer.  It was fairly exhausting work, but went pretty quickly.

Add water ...

We live in the country and have a well for water, but it is a poor quality well.  We can pump about 300 gallons from it before it runs dry, and it takes a few hours for it to replenish.  We need about 9,000 to 10,000 gallons for the deep zone alone, so filling it from our well is out-of-the-question.  Fortunately, there is a local water hauler who delivers 3000 gallons at a time at $110 a pop.  The delivered water is chlorinated, but by the time we are ready to turn on the pond pump and get this "natural" water filtration system going, that chlorine will be long gone.

We thought it would be a good idea to fill the deep zone initially to get as many liner wrinkles as feasible out of the swimming area.  The deep zone is the only area where people will be in contact with the liner (in the regeneration zone the liner is going to be covered with lots of gravel) so it should be made as good looking possible.  Also, the water will make sure that the liner is pressed against the walls and into the corners so we can establish exactly how much liner we have to drape over the outside of the deep zone walls and regeneration zone, and how much liner we need to splice on to finish the job.

It was amazing how fast the water ran out of the truck.  Before we knew it, there was already a couple of inches in the deep end and it was getting difficult to lift the liner and work the wrinkles from the center towards the edges.  We had to hustle to stay ahead of the water as it covered the floor, but we found that the wrinkles could be moved by lifting and pulling just a few inches at a time rather than lifting large section of liner.  Once the floor was done, smoothing the walls proceeded at a much more leisurely pace.  I was pleasantly surprised that, other than where we folded the liner to fit the corners, there is hardly a wrinkle the in the swim area -- I didn't think that was possible without a custom shaped liner.

Next jobs will be to start plumbing this thing and to splice in the extra width of liner. 

Local men prevail in contest with inanimate object

The day began with a series of surgical strikes by Murphy.  The liner arrived in a 12' wide roll wrapped with several layers of underlayment fabric to protect it in shipment.  Since the liner is 50' wide, it obviously had been folded in some pattern before it was rolled. With only our excavator Brian (God bless him) and I available to pull the 1200 pound liner over the pond we knew it was going to be a long day if we didn't get the liner positioned so that the folds worked in our favor.  After examining the ends of the roll, we had it all figured out and positioned the liner next to the pond accordingly!

The first thing we noticed upon removing the protective layers of fabric was a piece of paper taped to the liner that read "50' x 70'".  Due to my excellent short term memory capacity, I had clearly (but mistakenly) remembered ordering a 75' long liner and Brian had dutifully dug the pond to a size that would require all 75'.  Oh, well, nothing that an order for an additional 10' x 50' roll of liner can't fix!

This discovery was followed quickly by Brian's observation that a rodent of some sort had crawled into the rolled liner and had busied itself chewing a hole.  Apparently the rodent realized that the liner was folded and, therefore, chewed directly into the roll in an effort to cause as much damage as possible.  At this point, I was pretty sure that the pond project was well on its way to FUBAR status.  Mercifully, the biggest holes ended up being at the edges of the liner and would not be below the final water level. Only two small holes eventually had to be patched using the seaming tape I already had on hand.

Murphy's third stroke was revealed when we unrolled the liner (which was easier than expected) and found that the liner had been folded so that it could be positioned in the center of a pond and pulled outwards to the edges.  Despite our careful analysis, we had positioned the liner completely wrong, at the far edge of the pond.  The folds that were on top were such that we would have to pull 25' feet (i.e. 600 ponds) of the liner away from the pond before we could get to the folds that could be pulled towards the pond, and then we would have to pull the whole tamale (i.e., 1200 pounds) over the pond.  To my great relief, Brian devised a plan whereby we could work back and forth over the length of the liner, pulling just one 5' to 6' width at a time and always in the direction of the pond.  Despite a lot of extra walking, we soon got the hang of it and were able to work efficiently.  Even with several water breaks, we finished in under three hours.

A couple of other lessons learned were that black rubber in the sunshine gets really hot!  Also, wear gloves.  In addition to being heavy, the rubber clings very well to stuff and, after working just one section of the liner, the skin on the first knuckles of my fingers had been stripped off leaving raw flesh.  I hadn't noticed at first since I mistook the burning sensation as being due to the heat of the sun warmed rubber.

Next comes some water in the deep zone to help us smooth out all the wrinkles in the liner.

Underlayment

Even though we have a layer of sand in the pond, it's still a good idea to have some form of underlayment between the soil and the pond liner to protect the liner from punctures. Some recommend carpet remnants or even newspaper, and the web sites of the European natural pool designers call for "fleece" as an underlayment.

I bought the pond liner from an outfit on eBay (which I won't name or recommend here since their customer service was so lousy), who also offered a matching quantity of underlayment at a good price and free shipping, so I got that rather than scrounging remnants from the flooring outlet.  This type of product was also recommended by the liner manufacturer to ensure a full 25 year (or more) liner lifespan.  

As you can see from the photos, the underlayment is light green and very much like polar fleece or a heavy felt -- I assume this is the same material the Europeans refer to as pond fleece.  We also noticed that, except for color, the underlayment is almost identical to heavy duty filter fabric that is used in building roads, septic systems, foundation drainage systems, etc.  So, when we ran short of fleece due to overlapping and the irregular pond shape, I bought a roll filter fabric at the local plumbing and septic wholesaler.  The filter fabric comes in 300' rolls in a variety of widths.  A 6' x 300' roll was $108.

The underlayment is in and has been inspected by my helper.  Now it's either put the pond liner in or finish stacking the 7 cords of firewood that has been laying (or lying -- I can never remember which is correct) in the yard ... maybe I'll have lunch first.

Taking shape, looking cool

We went away for a week to spend some vacation time in Vacationland (i.e., Maine) and returned to find this masterpiece sculpted in soil.  Brian had been busy carving out the final pond shape, and backfilling and tamping the deep zone wall to about 2' of its top.  The deep zone got a fairly thin layer of sand which was also tamped into a nice, solid base.  He also dug a trench around the pond perimeter, which will be used to lock the pond liner in place (the edges of the liner drape into the trench, which is then backfilled with soil and tamped).  It is difficult to see in the photo, but Brian also cut lengths of scrap plastic drain tile and wrapped them around the wall braces that extend above the backfill.  The braces have fairly sharp edges and the drain tile wrapping should protect the pond liner from punctures or tears. 

Almost time to put the liner in -- a prospect that has been keeping me awake at night as the liner weighs 1200 pounds.  When it arrived on a semi a few weeks ago, the driver had no way to even move the liner to the rear of the trailer, not that it would have fit on the trailer's hydraulic drop gate even if he could have moved it.  Fortunately, I have a farm tractor with a front end loader and was able to tow the liner out of the truck with a set of chains and then carry it to the barn with the loader for storage.  How we are going to unroll the liner and stretch it over the pond remains a mystery.

Let's do this thing!

The shallow bowl shape of a traditional pond results in either a small deep zone in a moderate sized pond, or a moderate sized deep zone in a large pond.  To achieve a swimmable deep zone in a relatively small pond it helps to form the deep zone with vertical walls.  Two ways to do this are to (1) dig a deep zone, construct walls in it, excavate the surrounding regeneration zone and then lay a liner over the whole shebang, or (2) dig the entire pond, lay the liner in it and then construct a wood wall on top of the liner to separate the deep zone from the regeneration zone.  Each has its advantages and drawbacks.  For instance, option 1 is faster, but requires more liner and excavator work.  Option 2 would require less liner and excavator work for the same size pond,  but constructing custom walls on-site could be time consuming.

As you can see from the photos, we chose option 1.  I sourced galvanized steel in-ground pool walls from the local pool store.  At $3300 for a 16' x 28' structure with hardware, they were not cheap.  However, option 2 would have cost the same in materials suitable for long term submerged durability (white oak framing and Ipe decking to go over the framing), not to mention the labor involved.

With a nice, level excavated site, the walls were a snap to put together.  The walls are 42" tall in 4' or 8' lengths.  The corner sections are pre-formed in a 2' radius.  The panels are bolted together through pre-drilled holes and provided hardware -- a piece of cake!  Our excavator Brian and I had the wall together in less than three hours.

The next day we assembled and attached the provided steel bracing, one brace per panel seam.  Each brace is an isosceles triangle that extends out from the base of the wall to help withstand backpressure from the deep zone water and preserve the structural integrity of the pool.  The long leg of the braces use a rod and nut to help plumb the wall panels.

Once the wall is together, braced and plumb, it's time to pour a concrete collar around the outside to hold it in place permanently.  Concrete was poured from the bottom lip of the wall out to the ends of the braces (about 2') to a depth of 6" at a cost of $577.

Ephemeral pond design

An "ephemeral" design since, as a battle plan rarely survives first contact with the enemy, I anticipate that the pencil-on-paper pond design will not survive much beyond first contact with the reality of the hole in our side yard.   

Being illiterate in CAD, to the left is a basic design published by Oase on which I superimposed the proposed dimensions of our pond (click on it to see a larger image).  This design is not too far different from what we are planning.

Since you found this blog, I'll assume that you have some familiarity with the concept of a natural swimming pool.  Our pond will utilize a pump to draw water through four suction lines -- one from a skimmer and three from 4" diameter perforated flexible hose ("drain tile") buried  in the gravel of the regeneration zone.  The pump will discharge to two lines -- one to a waterfall and one to a discharge line in the deep zone.  Our dock and skimmer will be at the northeast corner of the pond (to take advantage of the prevailing wind from the southwest), while the waterfall and pump will be at the northwest corner.

Most of the regeneration zone will be about 9' wide and consist of natural cobble and gravel 2" to 4" in size.   The regeneration zone on the south side of the pond will be a large beach area of "micro-mini" pea gravel for kids to play in and to provide comfortable access to the deep zone.  Since the beach is larger in square footage, two lengths of drain tile will draw water through that area while one length of drain tile will be used elsewhere.

How did we size our pond?  Basically, we decided that a 45' pond would look nice in our side yard and that a 16' x 28' swim area would be adequate.  From the reading I have done, I believe the regeneration zone should have a square footage at least as large as the swim zone.  Our regeneration zone will be nearly three times the size of our swim zone so we should have plenty of excess filtering capacity.  Initially I planned on a smaller pond since the maximum EPDM pond liner width is 50', which effectively limits pond width to 30'.  Fortunately our excavator contractor Brian proposed that we cut 10' from my already purchased 50' x 75' liner and splice it to give a 60' width -- a great idea that will give us enough width for the beach area without having to shrink the size of our swim zone!

Coming soon ...

Thanks for checking out my blog!  In the near future I will begin presenting the sequence of construction of our swim pond, hopefully with enough detail to help others to gather some ideas about designing their own natural swimming pool and to make some judgement about what may or may not work.  I'm sure that, once we've finished, there will be lots of "I wish I would have" reflections ... hopefully none will be major.  I'll be giving specifics on materials used, suppliers, costs, and how things are put together.

In the meantime, above is a photo of the initial excavation.  The bare soil is the approximate overall pond size and the excavated hole in the center will be a 16' x 28' swim area.  The design was loosely based upon an image of a Bionova natural swimming pool I found on the web (see below).  To orient you, our house would be just beyond the right edge of the Bionova image, and is just beyond the lower edge of the first excavation photo above.  That photo was taken (and many of the photos on this blog will have been taken) from an attic window on the north side of our house.  More on the pond design in the next post.