Biological sand water filters and roof water harvesting

New information is available about rainwater harvesting from composition roofs. The slow sand filter described on slowsandfilter.org has been shown to remove petroleum hydrocarbons from water down to less than 1 part per ten million by weight; and this is when it is operating at its LEAST efficient rate of purification (at 32 degrees F). Also, petroleum hydrocarbons are present in roof water from the composition roof in the study in potentially harmful amounts if consumed over a long period of time (2.9 parts per million by weight). If a first flush diverter is used the, concentration is reduced to almost acceptable levels (less than 1 part per million by weight). The hydrocarbons may, in fact, be originating partially from local air pollution, because diesel fuel traces were found along with heavy oils. The implication here is that composition roofing is indeed a viable means of collecting rainwater, if proper filtering is used. A first flush diverter can be built easily with some pvc and recycled containers, and a slow sand filter can also be put together for minimal cost using recycled containers.

Another important fact uncovered is that rainwater from a roof may indeed vary WIDELY in the amount of Coliform bacteria present. The study has shown that water from a roof can contain about 60 cfu/100ml of coliform bacteria (relatively harmless) to 50,000 cfu/100ml (very likely to contain harmful pathogens and not safe to use), and that a slow sand filter will vary somewhat in its ability to remove pathogens; and also that a first flush diverter will remove a considerable number of the pathogens, but not enough to be safe. The slow sand filter has been shown repeatedly, by other studies (see the lit cited page on the above mentioned website) to remove coliform, ecoli, and fecal coliform to a safe level – from none detectable, in the best case, to 10 cfu/100ml in the worst case. In all cases the water from the filter is very non-turbid and should work well in conjunction with a UV filter
followed by an epa approved point of use tap water filter.

A biological sand filter (slow sand filter or Biosand filter) is maintained or cleaned by actively managing only the top 5 cm of sand. This top layer is either removed and replaced when the flow rate becomes unacceptable (typically 4 to 6 months) with clean sand or “wet harrowed” meaning the sand is gently agitated causing partial break up of the biological surface. A rapid sand filter is maintained and cleaned by forcefully injecting water backwards through the entire sand bed every 2 or 3 days. Backwashing a biological sand filter will destroy it and may result in people getting very sick if water from it is being consumed. A Frequently asked questions (FAQ) on biological sand filters can be found here.

The terminology used to describe water filters that uttilize sand and gravel to filter water can be confusing. A “slow sand filter” uses sand and biological methods to purify water. A “rapid sand filter” uses sand to filter water. Rapid sand filters do not make use of biological methods, although some biological action may take place between backwashing (cleaning). A “Biosand filter” is a modified slow sand filter, invented by Dr. David Manz. A Biosand filter uses biological and sand filtering and is designed to operate intermittantly. A “sand filter” is a general term used to refer to a water filter that uses sand as the filter media.

Effective rapid sand filtration requires that chemicals and coagulants be added to the water. The coagulants attract particles including some bacteria. When the water with these coagulated pieces flows through sand, the coagualted particles are physically blocked by the sand. The water that emerges must be treated with chemicals such as chlorine, or ozone because most of the harmful bacteria, viruses and protozoa are still in the water. A rapid sand filter fills up in several days and must be cleaned by “backwashing” – forcing air and water back through the sand with high pressure. This breaks loose the coagulated contaminants. The resulting muck must then be drained off and disposed of. This material is hazardous waste. The bacteria is not killed, it is contained. When the chlorine or ozone is added to the water from a rapid sand filter the water  is sterilized but not purified. Sterilized means the pathogens are inactivated, but their remains are still in the water. Purified means the pathogens are removed from the water.

A slow sand filter, and a Biosand filter work much differently than rapid sand filters. To explain this it is helpful to think about biology class. To purify means to remove, to sterilize means to inactivate. From a biological standpoint, rapid sand filter systems sterilize water, biosand filters and slow sand filters purify water. All water, with the exception of distilled water, will have organisms in it. The most pure water will have very few organisms. Most of these organisms are harmless to people. When water passes through sand due to only the force of gravity, these organisms, that live in oxygen rich water, will cling to the sand particles and start to feed on other organisms. In about 3 to 4 weeks a living layer of organisms forms in the top 5cm of sand. This layer continues to grow and will eventually spread throughout the filter with most of the activity concentrated in the top 5 to 10 cm of sand. Harmful organisms are “eaten” by this biological layer, which is often called the “Schmutzdecke” which is German for “dirt cover”. This results in purification. Eventually the layer on top gets so concentrated that it slows down the flow of water. To restore the flow of water, this biological layer must be “moderated” to allow more water to pass through at a higher rate. This is done in one of two ways. Either the top 2 cm of sand can be removed, or the surface can be “wet harrowed”. “Wet harrowing” means gently disturbing the biological layer and then allowing the resulting cloudy water to drain off. This is not backwashing. The lower layers of sand and organisms remain un-disturbed. Aside from biology, there are physical properties of a slow sand filter. As water and gravity act on the sand, a compaction and settling occur, which results in the sand becoming a “filter cake” which increases the effectiveness of the physical filtering properties of the system. Backwashing a slow sand filter will destroy this property and disrupt the biological action in the lower areas of the filter – in other words backwashing a slow sand filter or a biosand filter from the bottom of the filter will destroy it.  Perhaps the confusion on this method comes from the fact that some large slow sand filters use a form of backwashing to effectively wet harrow the filter. Water is forced through the top 20 centimeters ( 8 inches ) of the 4 foot deep sand bed to break loose the top layer of biological activity. The bottom 3 feet 4 inches is left completely undisturbed. This is not the same as what is used on a rapid sand filter. In a rapid sand filter the entire sand bed is forcefully disrupted regularly every 2 or 3 days. There is a huge difference here. Constructing a small slow sand filter, in a 5 gallon bucket or a 55 gallon barrel and then backwashing it to clean it is just plain stupid. If the water from this situation is being used for consumption, people will most certainly get very sick.

First; my humble apologies for being so slow to moderate comments. Vast improvement is in order. We have updated the software that runs the blog ( as of July 24 2009) which went perfectly. Moderating should be quicker now.

The slow sand filter works using biological processes. Water passes through a layer of sand and after about 3 weeks a layer of biological activity forms in the upper 2 to 3 centimeters of the sand. The bacteria in this layer sometimes called the “schmutzdecke” break down harmful bacteria such as “beaver fever cysts”, all coliform including the most deadly and the resulting water out of the filter is safe for consumption. A lively discussion about these filters would be very encouraging. Anyone who has comments, questions, or experience with these filters is encouraged to leave comments. The website http://www.shared-source-initiative.com/biosand_filter/biosand.html documents a year and a half of experience building and operating 2 of these filters built from mostly recycled materials. All tests so far have shown they work very well. Also the website http://www.slowsandfilter.org has more detailed information and an FAQ page.