The Inevitable First Questions

For serious koi-hobbyists, the very first question we never ask ourselves is “should we build a pond?” Leave that to those poor (or should we say fortunate) mortals who have yet to encounter koi, but who might have a yen for a water feature in their garden with lilies and goldfish. We probably all remember this phase?

More common first questions are “how large a pond can we fit into the available space”, or “ what will be the best design”, or “should we build in or out of the ground?” Although we should, we rarely at this stage ask ourselves can we either afford to build or run the pond! It is quite possible to work out approximate costs in advance and something I would strongly recommend at an early stage, for the simple reason that if you don’t, you are likely to be in for some nasty shocks along the way. Building and operating a koi pond can be an extremely expensive business. As a rough guide, which for Ian and I turned out to be remarkably accurate, an estimate of building costs is $1000 (around £500) per 1000 litres (approximately 222 gallons) of final water volume. This includes purchase of all fittings, as well as actual construction costs. Operating costs require you to estimate electricity used to run the pumps, gas to heat the water, actual water costs if you are on a meter and the cost of feeding and maintaining your koi. Not to mention actually buying the koi in the first place!

A Beginning

Having bought our home in West Pennant Hills, a suburb about fifteen miles north of central Sydney, Ian and I first intended to put an extremely overgrown and neglected quarter acre plot to rights before ever tackling a pond. However, life rarely works according to plan and koi disaster struck for us ten months later in that we lost the rented water space housing Ian’s original koi collection. Very good friends Gerard & Kazuyo McDonald, helped us out, making room in their pond for the koi, but it was imperative that we suspend garden refurbishment activity immediately and start digging! Gerard & Kazuyo could not be expected to house our koi for any longer than was absolutely necessary. Water space for any koi hobbyist is incredibly valuable and never more so than in Australia, where koi breeding is an intrinsic part of practically everyone’s hobby.

We had already earmarked two sites in the garden for ponds. The first; adjacent to the kitchen window and verandah area at the back of the house, will eventually become the main ornamental garden feature for large koi and is expected to be around 50,000 litres or 11,000 gallons. The second site is around the side of the house in a small courtyard area, which housed a vegetable plot with an orange and lemon tree, the wood shed and a paved area with a rotary washing line. Planned here was a smaller, formally shaped growing on pond for young koi. It was this second pond we decided to tackle first, although the main pond site is more accessible. However, we did get down to brass tacks eventually and decided that we just could not afford to build the main pond at that point in time!

The downside of this (very sensible) decision was a huge amount of hard manual labour for both of us. The site of the second pond is not accessible to digging equipment, therefore required excavation the old fashioned way. Pick, shovel, wheelbarrow, skip and a lot of blood, sweat and tears. It took the two of us over eight months to remove the earth, clay and eventually solid slate from an area 6.4 metres (21 feet) by 3.2 metres (10 feet 6 inches) to provide a final water depth of 2 metres (just over 6 feet). One of Ian’s favourite tales is that if had not asked me to stake out the dimensions, this pond would have been quite a bit smaller! In reality, the final 27,000 litre (6000 gallon) pond has proved to be a very useful size.

We always planned this pond to be partially out of the ground. On a sloping site, this has meant that at one end, the water level is 0.6m (2 feet) above ground and at the other it is 0.9m (3 feet). The resulting wall is nice to sit on, you can get closer to the koi and they are also easier to catch and lift out from a pond of this type. Another plus is that you don’t have to dig quite so deep, although you must remember to account for the extra depth bottom drains and concrete thickness requires when calculating when you can stop digging!

The Hole

The footprint of the pond is formally rectangular with one corner removed, because a passageway around the pond to the garden shed and back garden needed to be maintained. Generally, there are two approaches to pond excavation. If you are able to use a mechanical digger, a hole much larger than the pond is usually dug out and the pond constructed bottom up, within the resulting space. The second method is to construct a concrete collar to stabilise the earth, then dig down vertically. In Australia, in the Sydney area, the subsoil is generally heavy clay, shale or solid sandstone and with the dry conditions here, quite stable. Since most of our pond area was bordered by paving, we dispensed with the concrete collar. We never encountered any problem with wall collapse, even when we approached the limit of our excavation. The solid shale at the lower levels however, was appalling to dig out. Rock-like when dry, if dampened the surface turned to a disgusting greasy mud that clung to everything. We did get some rain during the digging months, although fortunately, on an infrequent basis. We attempted to keep the rain out of the excavation as much as possible, by rigging a taupaulin ‘roof’, but unfortunately, it was not very successful!

Digging operations were carried out in two stages, simply because it was easier to reach the wheelbarrow from a narrow trench. Therefore, we dug out a width of 1.5 metres over the full pond length first, closest to the fence. When we had reached the required depth, we then commenced on the second 1.5 metre strip. It was a wonderful day when we threw out the final shovel-full! Ian calculated we removed thirty cubic metres of ground-fill, by the time we completed the basic hole. Of course, that was just the beginning, although it took by far the longest time.

We next had to dig out a channel from the corner of the pond nearest the lemon tree, turn 90º around the back of the tree, rise half a metre and continue along the fence to the wood shed, the site for the settlement and filter chambers. Channels and spaces also had to be cut into the base of the pond to house the pipes from and the two bottom drains. The bottom drains, (one an air dome design and the other plain), were then fitted and 100mm pipe-work (4 inch) glued into position along the prepared channel. Before being concreted in, all the pipes were tested for leaks simply by adding a vertical section and filling them with water. Interestingly, the thick-walled pressure piping commonly used for UK koi ponds is not required in Australia; normal household sewer piping is substituted. Ian assured me it was quite strong enough. It was certainly much easier to handle, lighter in weight and simple to cut and fix into position.

Having concreted the drains and pipes into position at the bottom of the pond, a section of pipe was also set into the long wall adjacent to the paved area just below ground level. This was to become the deep-water return from the planned “Clarity” fractionator unit fed by a power surface skimmer.

Filtration Decisions

During this preparation period we also decided on and commissioned the filter and settlement chambers, after having seen two in successful operation on other ponds. Both chambers are fibreglass rectangles, interconnected by dual pipes. The settlement chamber, 1.95m (6’ 4”) by 1.27m (4’ 2”) by 0.66m (2’ 2”) has 2 baffles across its length made of a plastic fibrous material similar to Japanese filter matting, but more robust. These assist the settlement of solid waste, before the water progresses to the filter chamber. Because of the large bore pipes from the bottom drains and the length of the settlement area, the system, although very simple, is remarkable efficient, given a pond turnover of 20,000 litres an hour through it. Three standpipes fitted to the base of the settlement chamber facilitate removal of accumulated mulm, although to save water, we also siphon out solid waste. The second, slightly longer filter chamber, 2.37m (7’ 11”) by 1.27m (4’ 2”) by 0.66m (2’ 2”) is filled with Japanese matting cartridges laid over an ingenious tubular air delivery system. The result is that every square centimetre of the entire chamber is liberally supplied with air, maximising aerobic bacterial development and producing a very efficient biological filter. Ideally, the settlement chamber would be the same length as the filter chamber, but we were constrained by available space in the existing wood shed. We did not, at this stage, want to waste time and money reconstructing a perfectly serviceable shed.

Phase 2

During the wait for the filter and settlement chambers to be delivered, the next phase of construction got underway. Normally, when constructing a partially out of the ground pond, a foundation and block walls would be built to the required height. In Australia, where swimming pool builds are common, this technology is both preferable and cheaper. First, a template, or ‘form-work’ as it is called in Australia, is necessary for concrete to be sprayed against, which Ian built from timber supports plus 5ply plywood facing. On the inside of this and all over the rest of the pond walls and base a mesh of 12mm steel ‘rebar’ reinforcing rods was tied into place, supported by plastic support ‘bar chairs’ approximately 15cm (6”) above the surface. Ian assisted Lars Lindburg of AllCare Ponds in laying this framework and setting the surface skimmer box and all pipe-work, with eyeball directional inlets, in place. This included five 50mm (2”) returns from the filter, three on the end wall and two angled along one long side to ensure adequate water movement.

We now had to make sure everything was dry. Ian gave me the lovely job of mopping up every small puddle on the base of the pond from a recent rain shower – accomplished with a sponge in the end! It was the only tool I could think of to get inside the small areas around the bottom drains and in the pipe channel. It worked anyway, and we were then able to give the go ahead to the swimming pool construction crew who came and sprayed the entire structure with 11.2 cubic metres of concrete, covering all the steel, which acts as both a strengthening device as well as a support.

This construction method is both stronger and much faster than using a conventional block-work technique. Completed and smoothed off by steel trowel to a beautifully rounded finish in half a day, the installation also eliminates the need for rendering, forming an excellent koi pond when finally covered with fibreglass matting and painted with black gelcote. The outside, once the form-work was removed, was covered with light grey slate tiles.

Phase 3

We had now reached a critical stage. Gravity fed, the final water level in the settlement and filter chambers had to match that of the pond. Given the sloping ground, plus the above ground pond, this was no easy task. A very heavy-duty frame had to be constructed to raise the chambers to the required height, also allowing space for the drains below the chambers to be placed. Ian built the frame from old, but incredibly strong recycled hardwood in-filled with form-board. It was a massive construction, you would have thought it would support a Sherman tank, but in reality, we did end up with one major leak later on, because of about a 5mm give in the form-board in conjunction with a design fault in the settlement chamber. The baffle supports had been screwed into the fibreglass, as well as glued and a tiny hole developed. Luckily, an extra ‘leg’ underneath cured the problem, but it caused a lot of worry.

At this stage, we were waiting for deliver of the filter and settlement chambers. We purchased the pumps (two 92w, 12,000S Sequence plus one Speck Continental 90/500DS), capable of a 30,000 litre water turnover between them. Also two Hi-Blow 80 air pumps, a three tube UV assembly, a Clarity fractionation tower and all the remaining pipe-work, ball and gate valves necessary to complete the construction. A last minute addition was an interesting device called a multi cyclone, again from the swimming pool industry. It is essentially a small, but very efficient centrifuge, to act as a pre-filter for water to the fractionator.

We were then back to digging! A channel had to be constructed, from the surface skimmer, along the fence at the side of the house to the garden shed, the proposed site for the Speck pump servicing the Clarity unit. Additionally, we had to dig along the front edge of the pond wall to meet the deep-water return previously laid through the wall. This part of the digging operation was probably the toughest. Space was severely confined, the ground was full of rubble and we had to contend with existing drains, which we did not want to put holes through! However, the job was eventually accomplished, pipes laid; a foundation for the Clarity unit built and the pump and cyclone installed inside the shed. The out-flow from the Clarity was channelled into a 1.5m rectangular fibreglass box filled with crushed coral, designed to boost and stabilise pond pH and hardness. Ian enclosed the rather ugly box beautifully from spare blocks originally used for paving the area now taken up by the pond. From the box, water is piped quite steeply downhill and re-enters the pond. The entire system is fed by the power skimmer, powered by the Speck pump at approximately 10,000l / hour. I have to say; these Clarity fractionators seem to be incredibly efficient at removing protein or dissolved organic compounds (DOC) from the water, leaving it sparklingly clear and healthy looking. The amount of evil-smelling foam generated daily is pretty spectacular. The other claim however, that nitrate is substantially reduced, we have yet to observe. Perhaps this feature only develops with time, but to date, after running the pond for seven months; nitrate is still over 100.

Phase 4 – the end in sight!

At last! The filter and settlement chambers were delivered and installed by their designer, Ralph Böner. Ian assisted, but was very glad of the help at this stage. We were both so tired! Practically every weekend for over ten months, we had done nothing else but build this pond and maintain house and garden. It was a tough year and a very magical moment when we siphoned out the last wash-down water out of the pond, made sure everything was properly connected up and clean, attached the flow meter to the hose pipe and switched on! The system took about 24 hours to fill up, very slowly, to a total of 27,000 litres. Amazingly (to me, but not to Ian, who had every faith in his construction accuracy), the water level in the pond and settlement / filter chambers matched exactly! A dechlorination agent was added; pumps and air switched on and the system left to itself for two days. No leaks were discovered at this stage and everything appeared to be working perfectly. It was the middle of December by this time and the water temperature about 26ºC. Time to bring in the koi!

Normally, koi are introduced slowly, to a new pond, five or six only in the first few weeks. In our case, we could not do this, because Gerard & Kazuyo, having housed our koi for almost a year, now had acute water problems of their own. It was imperative to lighten the load of our koi on their system as quickly as possible. Over the year, we had sold quite a few at the bi-monthly koi auctions held in Sydney, but there were still between 40 and 50 koi to bring home. We introduced 20 first and the rest a week later, a huge load on a new system. We knew we were in for trouble, but we celebrated anyway and cracked a bottle of Champagne, sitting on the pond wall, watching the koi for the first time.

And Finally

Well, you don’t really need me to tell you what happened next! Massive ammonia overload and soaring pH. It takes about 3 months for a new filter to really begin to operate with any degree of efficiency, so those first few weeks were a very anxious time for us. We could not bring the ammonia down to safe levels by water changes, with so many koi involved, instead, we dropped the pH below 7.0, to keep the ammonia as non-toxic NH3, not NH4 which will kill the koi over an extended exposure at high levels (>0.5). I checked the ammonia and pH levels daily and controlled pH to approximately 6.7 by addition of concentrated hydrochloric acid. About 250 ml (half a pint) of acid was required every day, added very slowly to the exit water from the Clarity unit, since this returned to the pond in a large enough volume to dilute the acid quickly. We also filled the entire filter and settlement chamber surface area with water hyacinth, which directly utilise ammonia. Unfortunately, a side effect of this effort to assist reduction of ammonia was almost disastrous, because we introduced Trichodina with the plants. It took us a while to realise that the discomfort of the koi was due to something more than sub-standard water conditions. Once diagnosed from a skin scrape under the microscope we quickly controlled the parasite with salt. Very sadly however, we lost one of Ian’s old favourite koi, a Kin Ki Kokuryu. Irritated by the parasites, it jumped out one night and was beyond help, when found. Just a pond fish, but special.

Gradually, after 6 to 8 weeks, the filter began to operate and nitrite levels began to climb, while the ammonia dwindled. We lost no other koi during the entire running in period, which I think is quite an achievement. Water changes kept the nitrite levels acceptable and I allowed the pH to climb back to 7.3, once the ammonia became undetectable. As predicted, after 3 months, the nitrite also dropped to zero and we could bring the pH up to 7.6 with bicarbonate, at which level we now maintain the pond. All parameters have remained stable, although we add bicarbonate periodically to assist the hardness. Our water here is quite soft.

We went through a period when the water was very turbid, although not strictly ‘green’, because the UV breaks down algae cells. This lasted about a month between the end of March to the beginning of May. We have had no blanket weed and at all other times our water has looked spectacularly clear. The koi are doing well and have settled into their new home extremely well. Our only remaining worry is the high nitrate, which the Clarity unit has failed to reduce so far. Nitrate in itself is not harmful, but its presence indicates the water, however good it looks, is not quite as good as it should be. I guess we just keep working on it!

We are delighted with the results of our first largely DIY pond construction job. Just a bit of window dressing still to do, for example, we would like to tidy up the rather ugly fence with some bamboo. Then, on to pond number 2! Watch this space.