SWINE FARM WATER PURIFICATION AND RECYCLING

We were asked to design a recycling system for a hog farm where the existing waste water was too high in nitrates to be discharged to ground. The farm was in an area of heavy year round rain fall and the effluent had been dumped where it eventually went into a creek. The goal of the customer was to have clean water for washing down the hog barns and water for the animals to drink as well as a marketable product as solid waste fertilizer and a liquid concentrate with high nitrogen for the local sod farms and golf courses. Clean water in excess to needs and able to meet local water quality standards would be discharged to ground.

In this case, a primitive settling system already existed and the sludge was settled then the liquor was pumped to a holding tank where it was aerated and naturally occurring bacteria was allowed to react with the remain material for about 8 hours to reduce the nitrates. The solution was then pumped to a holding tank where it was then held for discharge. At this point we were asked to do something about the remaining high nitrate level and the high total dissolved solids level. The total daily flow was about 6-8,000 gallons but it was to be treated over an 16 hour period.

The water source was barn washdown water. The hog barns were washed down three times a day by an automatic timer. This resulted in peak flows that had to be leveled out. The batch holding ponds served to level out flow rates. The final holding tank had to be mostly emptied between cycles. The system designed and built consisted of an aluminum silicate media filter to prefilter the solution. Prior to contacting us, a previous attempt with a sand filter found the sand filter plugging in about 15 minutes of operation. The aluminum silicate media is easier to backwash and takes out larger solids without plugging.

The next stage was a 10 gpm ultrafilter (UF) operated at about 40 psi. This allowed us an easily cleanable stage and a single membrane to replace should the material completely plug the system. The ultrafilter concentrate was diverted back into first holding tank. It was planned to let the solids removed by the ultrafilter settle to the bottom of this tank and be drained off later. The feed for the prefilter/ultrafilter was located about 18 inches above the floor of the tank to prevent the solids from being recirculated back to the UF. Later this concentrate line was returned to the start of the settling system to allow more time for solids removal.

Initially, no polymer was used but it was found the TSS (total suspended solids) could be as high as 3% even after settling. This was too much for the system and a polymer was selected. Samples were sent out for testing and a polymer that could settle the material in about 30 minutes in a 1000 gallon tank was used. The cost for the polymer over a years time was estimated at about $700.00.

The permeate from the UF was sent to a tank for feeding the Reverse Osmosis system. The RO was a 10 gpm unit also. The feeds for all the units were microprocessor controlled and the systems were shut down automatically when no solution was available for processing.

The RO system was also plumbed in a recirculating mode and was left this way. The concentrate was fed back to the RO feed tank and a 1 gpm bleed was sent to the settling ponds for reprocessing. This maintained the RO feed TDS at about 2500 ppm. The permeate out of the RO system was about 20 ppm and crystal clear. The local well water was about 160 ppm.

No UV or other sterilization method was installed so the water was not used for feeding but it was recycled for wash water. Some clean water was sent to drain as there was a net water gain over a days operation. This system was in a very rainy area of the country so no water was need for irrigation.

During startup, it was found that if the system was allowed to set for more than two days, bacteria grew on the membranes and plugged them. Flows would drop. Cleaning with hydrogen peroxide solution killed the bacteria and high flow flushing restored the membranes to original flow rates.