Rain water harvesting and slow sand water filters

An update regarding our ongoing work on slow sand filters. All of the filters here are still working. The oldest, filter 1, is working after 8 continuous years running. Filters 2, 3, 4, and 5 are still working.

Here in Washington state, it has been a hot dry summer, and the fall is starting out dry also. If we get a repeat of last winter’s lack of snowfall retention in the Cascade mountains, there will be major concerns regarding water here in western Washington state.

On another similar topic; the Earth’s ground water is being used up faster than it is being replaced. In areas where there are large aquifers that are used to supply water, this is a huge problem.

It might be a good idea for people to learn about biological sand filtration and rain water harvesting sooner, rather than later. Think about water.  There’s lots of info in this blog about slow sand filtration and rain water harvesting.

A recent news article has warned of cryptosporidium parasites in pool water.  Another article warns of the same problem. These microbes cannot be removed or killed effectively by the use of chlorine. Ozone is effective but requires high concentrations of the chemical to kill cryptosporidium and the system must be designed specifically for that purpose.

There is an alternative that neither of these articles mentions. Slow sand filtration is the most effective way to actually remove cryptosporiduim from water.

From the EPA site on page 19:

“Of the technologies available to the drinking water industry, membrane processes (forms of micro- and ultra-filtration) appear to provide the most significant levels of Cryptosporidium removal. Conventional treatment practices appear capable of meeting 2-log removals in most of the cases studied to date. Although direct filtration and in-line filtration may be expected to be less effective than conventional treatment, this has not yet been demonstrated in a conclusive manner. Alternative technologies such as diatomaceous earth filtration and slow sand filtration appear capable of achieving comparable, if not better, levels of Cryptosporidium removal than conventional treatment. A comparison of removal efficiencies of some bench-, pilot-, and full-scale water treatment processes is presented in Table 3 below.”

Slow sand filters do not use  chemicals to filter water. They use naturally occurring beneficial non-toxic bacteria that actually eat cryptosporidium microbes and all other dangerous micobes. A slow sand filter is actually a small wetland that purifies water naturally. These filters are inexpensive and require very little maintenance and no added chemicals. They are totally sustainable. This entire blog is about slow sand filtration.  Why aren’t these people using them?

More about cryptosporidium removal here.

and here.

and here (see page 2 of the document) greater than 99.9 percent removal (removal not just killing – ozone or chlorine only kill the microbe)

Back in February of this year we wrote about the lack of snow in the Cascade mountains in Washington state. We hoped for more snow – that did not happen. Now the situation is worse. Let’s hope there is not an extended hot dry summer. Things will get very uncomfortable. For those who have set up rain barrels, the situation may not be as bad. Those who have 3 or 4 rain barrels and a slow sand filter running now will be in much better shape. It’s still not too late to set up a rain barrel or two. . . . .   .

From the Washington state department of Ecology’s website as of May 11: “24 of Washington’s river basins are in a declared drought emergency.”

Last year and the year before, we did some work on five gallon slow sand filters.

http://www.enlight-inc.com/blog/?p=1395

http://www.enlight-inc.com/blog/?p=1740

http://www.enlight-inc.com/blog/?p=1848

http://www.enlight-inc.com/blog/?p=1927

http://www.enlight-inc.com/blog/?p=2052

http://www.enlight-inc.com/blog/?p=2089

http://www.enlight-inc.com/blog/?p=2179

We set up 3 of them, ran them for a year and a half, and tested the output water. We found that they just barely worked at all. Several issues became apparent. First, the buckets are easily distorted by the weight of the sand. Because of this, if a filter bucket is moved or bumped; the biofilm will be broken and contaminated water will break through to the output. Second, the inside of a new plastic 5 gallon bucket is usually very smooth. This prevents a good seal from forming between the inside surface of the bucket and the sand. The result is that contaminated water slips down the side between the smooth surface of the bucket and the sand. Third; in order to facilitate effective filtration, an extremely slow flow rate must be maintained – much like that of a dripping faucet. Any faster, and the water does not have enough contact time at the top of the sand where the bacterial action takes place.

In the most dire of situations, one of these 5 gallon filters would be better than nothing if all the above issues are taken into consideration. Make sure the inside of the bucket is roughed up considerably, and make sure the flow rate is very, very ,very slow; about 1 litre per hour. Do not disturb the 5 gallon filter container once it is running. Use very fine sand, around .15 mm effective size. Be sure the water put into the filter is non-turbid. Muddy water will NOT work in one of these filters.  If you only have access to five gallon containers, consider using two 5 gallon filters in series. That is put water through one, and then through a second filter of the same size. If you use 2 filters like this, consider using a coarser sand in the first filter; around .35 mm effective size; and then .15 mm effective size in the second filter. Most importantly, know that this will work only marginally at best.

Chlorine can be removed from water by adding vitamin C to it. This could be a way to facilitate filtering tap water with a slow sand water filter. The chlorine in tap water makes using a biological sand water filter nearly impossible because the chlorine in the water will inhibit the growth of the beneficial bacteria that make a slow sand filter work. If the vitamin c inactivates the chlorine, then the water could be filtered with a slow sand filter.

Update, May 15:
44 percent of Washington state is in drought emergency condition From the department of ecology’s website: “24 of Washington’s river basins are in a declared drought emergency.” Some of our neighbors, who are on public water systems, have already received notices of water use restrictions.

Update; April 2:
The SWE (Snow Water Equivalent) is less than 50 percent for most of Washington state.

Update; March 25:
Drought has been declared in three areas

from the above website:
“Currently snowpack statewide is 26 percent of normal and the forecast calls for little to no snowfall into the spring and warmer-than-normal temperatures through the summer.”

—————————————————————————————————————————–

A large percentage of the snowpack in the Casacde mountains has melted. By Friday, February 13, the freezing level is forecast to be above 10,000 feet. Snow will melt. 

Rainfall totals are near average here for January and February to date, and temperatures have been above average for this time of year.  Unfortunately, this has caused lots of snow to melt in the Cascade mountains.  This means that unless we have some really cold weather and a lot more snow in the mountains, there may be water shortages this year in some areas. Those who have rain water harvesting systems set up will have plenty of water. People who depend on snow melting may have issues with water; it just depends on how much snow falls in the coming months in the mountains. There could be more cold weather and snow on the way for the mountains and that would help considerably. By next week at this time the forecast is for some cooling. Let’s hope it snows up there a lot. However, it is nearly the middle of February and enough snow to make up for what has been lost is not likely. If the summer gets hot and dry, and it may very well do just that, things could get quite uncomfortable. Be prepared, set up your rainwater harvesting systems now.

We found the 5 gallon slow sand filter design to be inferior to the 55 gallon design. The five gallon filters just don’t work as well. They are less efficient at removing biological contamination at best; and actually introduce contamination at worst.

Another study totally not related to ours found basically the same results (the need for 2 separate containers because 1 five gallon container was not enough to provide adequate filtering):

http://faculty.mercer.edu/lackey_l/documents/CompletedThesis.pdf

(see page 56 in the conclusion section)

When constructing a slow sand water filter, make sure the sand used is not contaminated.
Make sure the sand source provides the correct type of sand.
Running some chlorinated water through the filter before starting it running with unfiltered water is a good idea. If the sand is contaminated with bacteria, the filter may not produce clean water. Running five gallons of chlorinated water through the filter initially, will kill any residual bacteria present in the sand. Do not use chlorinated water after the first decontamination run, just the first time the filter starts, that’s it. After that the filter will build up a biolayer on the top of the sand where it is supposed to be.

Filters 1, 2, 3, 4, and 5 are still operating and producing clean water. The three pond filters are also still operating. Filter 1 was started in 2007.

We are working on a pdf that has the most usefull information from all three websites and this blog summarized in one document. This will be available for download sometime in the near future (2 or 3 months).

Only two of the five gallon filters are still running. The others have been removed from use because they just do not work dependably enough. The two five gallon filters still running are set up in series. Water from the first flows directly into the second. So far, it looks like this set up may work; however the flow rate is very, very slow; not much more than 1 litre per hour.