Slow sand filter construction: a brief summary of success and failure

This is a summary of 6 years of experience building operating and testing small slow sand water filters. There have been a significant number of failures which have been good learning experiences, subsequently resulting in astounding successes. The failures are listed with the solution resulting in success as the “Fix”. Eight different large units have been built modified, studied; and all are currently in operation. Each of the filters are slightly different but all based on the same operating principles. One experimental small unit is also operating indoors on the counter top. The larger units are all built with mostly recycled containers made of food grade plastic. Sixty five gallon, 45 gallon and 55 gallon capacities have been used.

Failures and successes have been due to the following:

1. Sand size too large and not evenly sized:”construction” grade sand does not work well – coliform not removed completely:

Fix: The sand must be at least .25 mm effective size with a uniformity coefficient of less than 2. (the sand must be evenly sized and fine grained).  The best combination is .15 mm effective size on top half and .25 mm effective size on the bottom half.

2. Freezing weather; if the filter freezes it will cease to function and pipe damage may occur.

Fix: Keep the filters from freezing solid. Here, in western Washington state this really is not a big problem, however there are places where freezing will destroy a small sand filter. Several weeks of below 20 degree weather is survivable if the filter is kept running with water supplied from a non-freezing source. Several months of below 10 degree weather might ruin the filter.

3. Highly turbid water will clog the filter very quickly, all of the filters running on roofwater have at one time or another become clogged resulting in very slow output flow.

Fix: Supply non turbid water to the filter input. A pre-filter or first flush diverter is necessary. If the water is above 10 ntu’s it will drastically slow down the flow.

4. Air pockets forming in the sand during construction, adding water to barrel full of sand is a guarantee for disaster. This happened once here Never again. DO NOT JUST PUT SAND IN THE BARREL AND THEN ADD WATER. This will not work and you will end up with smelly toxic water output: that happened here and 1000 pounds of sand had to be shoveled out – wet and smelly – and completely cleaned and then shoveled back in.

Fix: When the filter is built, add the water first; then put the sand in slowly stirring it to be sure all air bubbles are gone.

5. Insufficient drain system.

Fix: Make sure there is adequate drainage at the bottom of the inside of the barrel under the gravel; a good size is quarter inch holes drilled in 1/2 inch pvc pipes with at least 6 pipes covering the entire bottom of the barrel. Put the drain system in and make sure the flow out is just as if it was flowing out of a 1/2 inch hole in the bottom of the barrel.

6. Sand not washed thoroughly enough causing cloudy water output.

Fix: Wash the sand until the water flowing from it is CLEAR. If you don’t it will take several months or more for the water to clear up.

7. The use of a barrel that held onions.

Fix: The most recent filter put together here used a barrel that had held onions. That odor does not go away and just gets worse. The water output is clear and most likely bacteria free, but very undesirable until the water sits for several days. If you have a barrel that had onions in it make sure you clean it out somehow before using it. I don’t know how to get that odor out. Update July 17 2012: The filter output is now free of the onion odor. I have been running water through the filter manually for the past week at maximum flow, apparently enough water through the filter will slowly remove the odor. Water dissolves most things quite well.

8. Improper first flush diverter function causing turbid water to clog the filter.

Fix: Make sure the first flush divereter is functional. The divereter here was neglected for several weeks and the filter clogged up from non-turbid water spilling over from the diverter. Make sure the “slow drain” is on and functional so the diverter empties after each rain event.

The small unit in the house is just that – very small. A 2 litre clear plastic pop bottle with the top 2 inches cut off is filled with 5 inches of  .15 mm effective size sand with 1 inch of small gravel (1/4 inch diameter pebbles) in the bottom. A plastic straw is used as the output pipe sealed with plumbers putty at the output located at the bottom of the bottle. A 1 quart plastic yogurt container with small holes in the bottom is used as the baffle – it fits snugly into the bottle. The only cost was the straw and the plumbers putty, and the sand at the most 25 cents. The rest is recycled.  We use it to filter well water. So far it is working and shows the schmutzdecke forming at the top. The water out is clearer and tastes better than the water straight from the faucet. I never thought this could work. . . .  hmmmm. Tests will go in when time and money allow.

This entry was posted in slow sand water filter study and construction. Bookmark the permalink.

2 Responses to Slow sand filter construction: a brief summary of success and failure

  1. dan breeding says:

    I am in process of constructing sand filter. I understand my priorites should be:
    1) have uniformed size particles for the top portion of the sand filter.
    2) the smaller the particle the more effective the operation of the sand filter.
    3) the smaller the particle, the slower the water flow.
    4) To avoid find sand ending up with the good water, a layer coarse sand should placed on top of the gravel, but the depth of sand and gravel is not important as long as the the fine sand does not reach the collection manifold.
    Is statement 4 correct?


  2. Orpheus says:

    In my humble opinion, statement four is mostly correct as a general guideline; based on the experience with filters at this location – however; with some added info:

    All of the successfully tested filters here have at least 22 inches of sand above the gravel (including the coarse sand above the gravel) and are contained in 55gallon, 45 gallon, or 65 gallon recycled food grade containers. The depth of the coarse sand is not absolutely critical – it can be anywhere from 6 inches to half of the total sand in the filter (assuming at least 22 inches of sand total and .45 mm effective size for the coarse sand). Have at least 5 inches (preferably 8 – 10 inches) of space for input water over the top of the fine sand.

    Assuming the filter container is a 55 gallon barrel, or a 65 gallon, or a 45 gallon barrel; on the bottom, the gravel should cover the drainpipes by at least 1 inch but not more than 3 inches. The coarse sand that sits directly on top of that gravel should be enough to keep the fine sand from working its way down through the gravel. In the filters operating here, that coarse sand is not more than half of the depth of the total depth of sand in the filter, but not less than 6 inches.

    It is also very important to make sure the sand and gravel are thoroughly washed. The water should flow absolutely clear when flowing from the washed the sand and/or gravel before using it.

    There is one very small filter in operation here that only has 5 inches of sand above the gravel; however, it has not been tested and uses very non-turbid water as a source. If anyone has experience with a small filter like this please don’t hesitate to post your test results and working experience notes here for others to read.

Leave a Reply

Your email address will not be published. Required fields are marked *

thirty three − 31 =