My homemade reedbed water harvesting project
||Grey water reclamation at home
||Me / Planet Earth
|Cost / Value:
||Spring 2006 – Spring 2007
||How to beat the hosepipe ban and save the planet
|The inspiration and research:
||Article on Monday, 13 March 2006
Article on Monday, 13 March 2006, extracted from BBC News.
Britain's biggest water company will ban hosepipes and sprinklers from next month, the firm has announced. Thames Water, whose eight million customers will be affected by the ban, says two unusually dry winters have caused "serious" water shortages. The South East has experienced its driest period for more than 80 years
I used the water usage calculator on the BBC website and estimated that my family uses approximately 400lts of water per day. That equates to 133lts per person which compares favourably to the national average of 155lts per day. Of this about 60% is for showers and baths. However, the above calculation did not include the irrigation system that I have in the garden. I recalculated the water usage including use of a hosepipe to water the garden, but for a reduced time to account for the difference between a hose pipe and irrigation system rate of flow. This added 90 lts per day. Coincidently the water used for showers and baths equates to 80 lts per day.
We already did some water conservation by collecting rain water using a single water butt and leaving the grass to grow a little longer, and not watering it. So it seemed a simple solution to use the bath waste water to water the garden. At this point I should have just decided to have showers with the plug in and siphon out the water with a hose after it had cooled. As sane people eventually did. You can even now buy special products specifically designed to make it easier.
However, I decided to research the possibility of creating a grey water reclamation system using reed beds. The local library proved, yet again, to be a good source of information, as did the internet. I learnt,
- that different reeds deal with different pollutants and pathogens,
- about aerobic and anaerobic digestion
- micro-organisms, bacteria, fungi, and protozoa
- surface or subsurface, horizontal or vertical flow configuration
- rainfall patterns throughout the year
- rate of flow through reed beds
- and area of reed bed required per person
- Pollutants and pathogens are removed from the waste water flowing through a reed bed by a complex variety of physical, chemical and biological processes, including aerobic and anaerobic microbial activity, nitrification, plant uptake, sedimentation, precipitation and filtration.
- Reed beds are successfully and economically used in full scale black water sewage treatment and industrial effluent treatment.
Importantly reed beds require little maintenance, no additional chemicals, little or no energy dependent upon the site geography / topography and are of course, completely natural.
I wanted the system to be low maintenance and durable. It also had to be aesthetically pleasing, sustainable and capable of providing adequate clean water supplies during the increased period of drought that may be experienced due to global warming. The site is relatively flat with only a slight fall towards the top of Figure 1 ‘The Site before start’ below. To minimise the energy requirement of the system as much as possible, it had to be gravity fed. However, it ultimately had to be landscaped into the garden so some pumping was going to be necessary.
The previous research indicated the there were two basic configurations, surface and sub-surface flow, with the later being further divided into, horizontal flow and vertical flow. The sub-surface configuration is thought to be better for temperate climates, especially during winter, as the water flows through the substrate, thereby staying warmer and more efficient. To have a sub-surface vertical flow configuration would involve having outlets at the bottom of the tubs, which would necessarily involve burying pipe work, access chambers and making holes in the bottom of an otherwise water tight vessel. It would also require a relatively complex distribution system for the grey water. A sub-surface horizontal system has similar difficulties with respect to the outlet but without the inlet difficulties. I considered that the sub-surface configuration imported too much risk of failure overtime and the buried pipe work and access chambers unnecessarily complicated for the relatively small gains offered by sub-surface flow configuration compared to surface flow. The surface flow configuration has the water flow above the substrate, through the reeds, and is more similar to natural wetlands. The reduced efficiency during winter can be countered by having a larger area, together with the ability to switch to normal direct discharge into the sewers if required. Therefore the surface flow configuration was adopted.
I also considered having all of the main tubs at exactly the same level so as to provide the infinity pool type look, but decided that this was impracticable, and less attractive than gently flowing water from one tub to the next, until quietly disappearing underground.
Another significant design criteria was that the system had to be operational, at least in part, in the shortest possible time and preferably before the hose pipe ban came into force. The project was therefore done in distinct phases.
The photograph Figure 1 ‘The Site before start’ below, shows the hose pipe being put to good use, marking out the edge of the development. Several shapes and locations within the garden were considered using this technique. This is the final location but not the ultimate layout. This also represents the ‘before’ photo.
Figure 1 ‘The Site before start’The location needed to be relatively close to the house for piping the waste water from the bathroom to the surge tank. It is necessary to have a surge tank to capture a bath full of waste water, and then to allow it to flow at a regulated speed through the system so as to allow the natural processing to take place. Consideration also had to be given for the wellbeing of the plants, such that they would not always be in the shadow of the house.
Another consideration was what to do with the excavated material. I did not want the normal hole and hill, nor did I want to have to cart the material off site. Adjacent to the site, also near the house, there is an old brick built shed.
The solution adopted was to increase the thermal mass and insulation of the shed by creating a turf wall near the shed and back filling the intervening space with the sub-soil excavated. The topsoil was stockpiled for use elsewhere in the garden. The shed wall was protected with two layers of waterproof membrane to avoid damp penetration. In turn the waterproof membrane was protected by reused expanded cardboard to reduce the risk of puncture by broken flints, primarily during construction and settlement. This additional wall also incorporated an old fashioned cistern arrangement which captures all of the rain water from the shed into open water. This overflows into the reed bed system.
It is planned that eventually the shed will also be re-roofed with a living roof. Combined this will provide a much reduced visual impact to the shed together with the benefits of a larger planted area.
The pumps necessary for irrigation circulation and to lift the reclaimed water from the sump pump chamber, the lowest part of the system, to the water storage are electric. In addition to this there is a small 12v fountain in the lowest tub, which now contains fish, and a pump to power the waterfall. Both the waterfall and the fountain are primarily for aesthetics, visual and sound, but they also provide additional aeration and agitation as a by-product. The electricity for the pumps is generally provided by two 18w photoelectric panels feeding a 12v 110Ah deep cycle leisure battery. The 12v supply from the battery is converted to the required 240v by a Sterling 600W inverter. The sump pump has to be connected to the mains to avoid the continuous use of electricity required by the standing current of the inverter if it was left on all of the time. The sump pump operates automatically by float switch which activates as soon as the sump pump chamber nears capacity and therefore requires a constant supply. The electrical requirement for the sump pump is however offset by the reduction in both the water requirement and the sewerage processing, which are intensive power consumers. Hence, despite not being fully power self sufficient, it is better than carbon neutral, it has reduced our overall carbon footprint.
The rate of flow required through the system has to be slow enough to process the water but not so slow for the water to become stagnant. It also has to be fast enough to deal with the input of grey water on a daily basis and ultimately to provide sufficient surplus to provide adequate stored water for later use. The research revealed that 1-2 square metres of reed beds are required per person to process black water but not the rate of flow through the system for grey water.
The solution was to experiment within phase one of the project.
As previously stated, pollutants and pathogens are removed from the waste water flowing through a reed bed by a complex variety of physical, chemical and biological processes, including aerobic and anaerobic microbial activity, nitrification, plant uptake, sedimentation, precipitation and filtration
The waste water treatment is outlined below;
- suspended solids settle to the bottom in still water or are filtered by the substrates and plants
- organic material is broken down by microbes that live on the roots and rhizomes
- nitrates can be taken up by the plants, or they can be transformed by denitrifying bacteria to nitrogen gas
- ammonia is transformed by bacteria to nitrates
- phosphorus precipitates with calcium, iron and aluminium compounds and is subsequently removed by sedimentation and absorption to the soil and by plant uptake
- metals and toxic chemicals are removed by oxidation, precipitation and plant uptake
- pathogens die off in inhospitable environment and are ingested by other organisms, or are killed off by antibacterial compounds.
The desktop research indicated which plants would be suitable in terms of their ability to deal with different pollutants and pathogens and to process the grey / black waste water. The environment required to carry out the water treatment outlined above could be achieved within a small site with careful design, construction and selection of plants.
Additional requirements that I wanted included, that the mixed varieties should be aesthetically pleasing, readily available, manageable, some flowering, indigenous, generally of UK origin and hardy. Such a mix of plants would inevitably have different size, spread and rates of growth. One of the listed plants is bulrushes. If they were not constrained they would quickly overrun most of the other plants. The use of separate tubs, and careful selection of which plants share tubs should eradicate this problem.
The design development and phase one of construction involved commissioning one tub only and testing the plants ability to process the water.
Figure 2 – ‘Phase 1 – The first filling’The first tub, with a water capacity before planting of 370lts, was placed in the excavation on a thin bed of sharp sand. Layers of large and medium sized stones recovered from the excavation were placed into the tub followed by a bag of pea shingle. This was topped off by a layer of topsoil, previously stockpiled, as the primary growing medium for the plants. This reduced the water capacity by approximately 150lts. The waste pipe from the bath / shower was cut and a low pressure water switch installed. One arm was returned to the waste stack and the other lead to the first tub, later the surge tank. Joints and fittings were checked and the test system declared ready.
Figure 3 – ‘Phase 1 – The first filling’Figures 2 & 3 – ‘Phase 1 – The first filling’ shows the first tub just after the first full filling. The water capacity at this stage represented bath / showers for the whole family. A mixture of reeds, flags and water hyacinth was obtained from the local garden centre and planted as show in Figure 4 – ‘Phase 1 – Planted, landscaped and operational’.
Figure 4 – ‘Phase 1 - Planted, landscaped and operational’The density of initial planting was relatively low so as to allow for growth, and to reduce costs in plant acquisition. The suction pipe of the irrigation circulation pump can be seen lying adjacent to the white waste water pipe which supplies the tub. Grey water provided in the morning was distributed in the evening which emulated a tidal effect. The plants in the tub quickly established themselves providing daily clear water fit for irrigation. The trial was considered a success.
Figure 5 – ‘Phase 1 - Planted, landscaped and operational’Grey water provided in the morning clears quickly during the day.
ollowing the completion of the trial more tubs were procured, again from the local garden centre. I decided that three additional tubs would provide adequate capacity and an appropriate rate of flow through the system. A larger surface area final tub would provide an interim holding area and centre of focus for the subsequent landscaping. The centre of this final tub was positioned and marked with a cane which became the setting out point for the other tubs.
Then the digging in earnest could commence. By this time the hose pipe ban had been in place for some time and the ground had become rock hard. A pick axe was required even at the sub-soil level. The sub-soil layer slowly gave way to clay and chalk and then to solid hard clay.
Figure 7 – ‘Holes slowly getting deeper in hard ground’The lowest part of the dig for the final tub was just at ground water level. Any deeper and floatation problems would have had to been considered.
The second tub has a rim height which allows the full body of water above the substrata of the first tub to flow into it. The rate of flow for the whole system is set by a simple pond hose tap on the outlet from the first tub. Each of the tubs is slightly lower that the preceding tub. A 1” pond hose is connected to the upper tub and overflows into the lower tub across the rim. This provides separation of the waters of the subsequent tubs and the slight sound of moving water.
The final tub overflows the rim into a drain channel and is collected into a sump pump chamber for later collection or distribution.
Figure 8 – ‘All tubs installed and landscaping commenced with a waterfall’Figure 8 – ‘All tubs installed and landscaping commenced with a waterfall’ shows an aerial view of the system complete and operational. The water hyacinths have grown remarkably well and make it difficult to see the tubs and water beneath them. The final tub can be seen more clearly with the metal cover of the drain channel showing adjacent to it. At the end of the metal cover there is the black cover of the sump pump chamber. This is in fact a 4 gallon expansion tank. This is an example of the utilisation of materials produced for different purposes being used for this project. This is the case for most of the materials other that aggregates and reeds. Sourcing some of the requirements proved to be an exercise in lateral thought and was something of a challenge.
Figure 9 – ‘Winter, and early morning snowfall slows progress’Winter set in before the landscaping was finished and effectively stopped work until spring 2007.
This proved to be beneficial. It provided sufficient time to identify two changes that I wanted to make prior to finishing the landscaping.
Figure 10 – ‘Redesign required, bigger pump sump dug’The sump pump chamber needed to be enlarged to accommodate a larger sump pump. The original was a clear water submersible pump with a low clearance so as to be able to pump to within 3mm of the floor. However, even fine sediment in the sump could jab the impeller. This increased the maintenance requirements considerably, to an unacceptable level. I decided to replace the sump pump with a submersible pump which was of a different design that was capable of handling solids up to 30mm. This was therefore unlikely to get jammed by fine sediment or even small stones. Unfortunately the replacement pump was physically significantly bigger that the first, especially in height. Therefore the sump pump chamber had to be taller, deeper and bigger capacity as the second also had a higher rate of flow. This is shown in Figure 10 – ‘Redesign required, bigger pump sump dug’.
Figure 11 – ‘Redesign required, bigger pump sump dug’At the same time as replacing the sump pump chamber I decided to add a ground source heat exchanger. It is partly experimental and partly to provide sufficient heat to the lower tub to prevent it freezing over totally, thereby allowing an ice hole so that the fish could breath. Figure 12 – ‘Ground source heat exchanger being prepared’, below shows it just prior to being lowered into the excavation. The deeper excavation for the sump pump chamber and the heat exchanger, actually a central heating radiator, meant that the heat exchanger was over 1.2 m deep and within both the summer and winter water table. The depth alone should be sufficient to not be effected by frost.
Figure 12 – ‘Ground source heat exchanger being prepared’In this test the circulation is gravity feed. The supply pipes are insulated and the heat transporting fluid is antifreeze as opposed to water so that it does not freeze at the surface. The pipes are coiled at the lower tub water surface to provide the heating element. Provision is made in the installation to be able to check for leaks and maintain fluid levels and for a post fit circulation pump if the gravity circulation system proves to be too inefficient. If required the circulation pump will be powered by the solar battery described above. The assessment of the effectiveness of the ground source heat exchanger will have to wait until next winter.
Figure 13 – ‘Ground source heat exchanger being lowered into hole’
Figure 14 – ‘Ground source heat exchanger beneath pump sump’
Figure 15 – ‘New Surge tank under construction - first layer’
Figure 16 – ‘New Surge tank under construction - second layer’
Figure 17 – ‘New Surge tank under construction - third layer (topsoil) and first planting’
Figure 18 – ‘Landscaping nearing completion, viewed from the pier side’The hard landscaping is procured locally but unfortunately sourced from the lake district. A case of aesthetics over green. The rocks are 1 tonne of Lakeland Green rockery and 1 tonne of Lakeland Green stone walling. So maybe it is green after all.
The pier is decking, just large enough to place a deck chair on, and do some fishing? It is actually the cover for the sump pump chamber and is easily removable for maintenance.
Figure 19 – ‘Landscaping nearing completion, viewed from the house side’, Figure 18 – ‘Landscaping nearing completion, viewed from the pier side’, above, and, Figure 19 – ‘Landscaping nearing completion, viewed from the house side’, right, shows the aquatic planting complete.
As previously stated, different reeds deal with different pollutants and pathogens. Certain plants, such as common reeds (Phragmites australis) and cattails (Typha spp.), have hollow stems that can transport air to the roots, supplying microbes with additional oxygen. Some take up specific metals or chemicals, other produce an exudate that kills pathogens.
Some of the plants in the system are listed below together with their ‘special abilities’.
Common reeds (Phragmites australis) and Cattails (Typha spp.)
flocculate colloids, eliminate pathogens
Bulrushes (Schoenoplectus spp.)
take up copper, cobalt, nickel, manganese, chlorinated hydrocarbons
Grasses (Scirpus spp.)
break down phenols
Rushes (Juncus spp.)
treat chlorinated hydrocarbons, cyanide compounds, phenols
Yellow Flag (Iris pseudocorus)
The water treatment provided by the above plants is supplemented by floating aquatic plants such as Water Hyacinth and Duckweed. Unfortunately Water Hyacinths are not hardy. However, they can be readily replaced with a few plants from the garden centre during spring. They are very prolific and will soon cover the tubs again. Alternatively a few plants could be over wintered in a frost free environment. Duckweeds are one of the smallest flowering plants and have one of the fastest reproduction rates.
Water Hyacinths (Eichornia crassipes) and Duckweeds (Lemna, Spirodella and Wolffia)
Take up nitrogen and phosphorus
Microbes living on roots transform nitrogen to ammonia
Take up trace metals, Boron, Copper, Iron, Manganese, Lead, Cadmium, Chromium Arsenic
The grey water from the bath / shower flows through this assortment of plants to produce crystal clear water.
The final planting is alpines and herbs in amongst the rocks, producing the desired ultimate effect of a small water treatment plant fully integrated into a garden environment.
Figure 20 – ‘Landscaping and additional planting complete, viewed from the pier side’
Figure 21 – ‘Landscaping and additional planting complete, viewed from above’The following photographs show the finished system.
Figure 22 – ‘David preparing to release fish.’
Figure 23 – ‘Landscaping and additional planting complete, viewed from the house side’
Figure 24 – ‘The waterfall’
Figure 25 – ‘Bulrushes in the primary surge tank’The surge tank has to be large enough to accommodate all the family having baths / showers in succession and to allow for the increased density of plants envisaged in future years. The plants and their roots and rhizomes, as they grow, significantly reduce the available volume for waste water.
arlier this year, just in one tub, covered in duckweed, nine frogs were counted. Newts have also colonized the system.
The photographs below are just an indication of the wildlife that lives or visits.
Figure 27 – ‘Wildlife – Common Newt (Triturus vulgaris)’
Figure 28 – ‘Wildlife – Common Newt (Triturus vulgaris)’
Figure 29 – ‘Wildlife – Common Newt (Triturus vulgaris)’
Figure 30 – ‘Wildlife - Red Mason Bee (Osmia Rufa)’
Figure 31 – ‘Wildlife – Large Red Damselfly (Pyrrhosoma nymphula)’
Figure 32 – ‘Wildlife – Common Blue Damselfly (Enallagma cyathigerum)’
Figure 33 – ‘Wildlife – Two Large Red Damselfly mating (Pyrrhosoma nymphula)’
Figure 34 – ‘Wildlife – Two Large Red Damselfly mating (Pyrrhosoma nymphula)’
Figure 35 – ‘Bulrushes (Typha latifolia) in summer’
Figure 36 – ‘Bulrushes (Typha latifolia) in spring’Figure 37 – ‘Bulrushes (Typha latifolia) in spring’
The excess water reclaimed needs to be stored and there needs to be a reserve of stored water in times of drought and / or hot weather to supplement the daily production of reclaimed water. At the start of this paper it was established that I needed approximately 90 lts a day to irrigate the garden at the same intensity as previously done. If I required to store adequate water for just half a year’s supply I would have to produce and store in excess of 16,000 lts of water. This is a huge container. However, when consideration is given to the rainfall pattern in Southern England the storage requirement is significantly reduced. If I were only relying on rainfall as my supply of usable water I would have to provide storage for 1,700 lts. This would be amassed over winter and alternately used and partially replenished during summer. With typical rainfall patterns the reserve of water would be almost depleted by the end of summer, relying on autumn rains to cope with an India Summer. However, as this cycle is supplemented by the reclamation process I have decided to limit my storage ability to 800 lts. This is provided by a cluster of four domestic water buts linked together. Other storage options considered included, a pond, a reused orange juice container, of the shipping variety – 1,700 lts or an ISO liquid container – 1,000 lts.
The sump pump fills the water buts and the irrigation pump draws water as required. If there is excess water it will overflow into a small area of native trees planted at the bottom of the garden. The amount of excess water is not expected to be as large as may be expected by calculation of requirements against bath / shower water usage. The reclamation process has inevitable losses. The aquatic plants have a high water take up. Surface water evaporation is significant in hot weather. Some of the water loss is designed into the system by providing direct systems of water draw into areas adjacent to the tubs to provide continuously damp areas for moisture loving plants, so as to enhance the natural look of the system.
In conclusion this has been a very rewarding project involving a number of different ‘green’ techniques to provide a solution to a specific problem. Albeit that this problem has yet to occur this year, the system is still working very well, producing crystal clear water day after day, whatever the weather. I have not had the water tested to be able to categorically state its quality, but suspect that it is actually potable. I believe that it has significantly reduced the family’s carbon footprint and provided interest and ongoing enjoyment.
I have added a narrow water but at the front of the house to capture rainfall from those elevations. The water but is small and relatively unobtrusive. It is however sufficient to be a surge tank to accommodate the build-up during downpours which is free to flow via a garden hosepipe to the reed beds in the back garden. The rain water flush provide by the front water but and the brick shed cistern also reduce the systems maintenance requirement.
Another development to be considered is the adaptation of the WC supply to be both mains and reclaimed water. This would further reduce our mains water consumption.
Finally my thanks to the books and internet articles that provided the invaluable reference material that made this project possible.