The Digester: Manure Pump

So here's an engineering challenge for you. Go find some oatmeal, and make it up nice and thick. Take a cup and a half of sand and pour it into the oatmeal, and mix it real good. Next, add in a handfill of grass clippings, some pebbles, and a piece of a chewed up leather glove just for effect. Now, fill a 100 square foot pit 12 feet deep with the stuff, and try and pump it. This mixture is the closest I can imagine to what sand laden dairy manure is like without talking about viscosity and thixotropic behaviors. Being a slurry of materials designed by nature to be difficult to deal with due to its sand content, sand laden manure has the ability to destroy pumps, tractors, and friendships with reckless abandon.

Farms are traditionally equipped with large, tractor-driven pumps to move this material between collection pits and trucks for hauling. Pictures of manure pumping equipment made by a particular manufacturer can be seen here . These pumps tend to be a large maintenance problem that require frequent replacement and maintenance, primarily because the impeller or shaft seals gets destroyed by the sand. Our original solution was a Moyno progressive cavity pump, which we feel is ultimately the way to go, but this too requires regular maintenance as the rubber bits inside wear out. Guillaume, our resident Crazy Frenchman suggested using a vacuum pump. It's a research project! Sounds great. That's how I got sucked into helping design a new type of manure pump, and ultimately getting my first faceful of deliciously crunchy sand laden dairy manure.

2 Inch PVC pipe clogged with sand An image of one of the pipes in our system. This is a 1.5 inch pipe, and is nearly half-filled with sand that had settled out of the manure. One lesson about pumping this material: As soon as it stops flowing, everything begins to settle. Even if the flow is continuous, you still get puddling of material around steps in pipes, like at glue fittings and valve seats.

Unsuccessful filter screen It's not just sand, of course, but also a lot of fibrous material as well. When we were initially trying to get our too-small Moyno to work, we tried putting a pre screen in the suction line to prevent the pump from ingesting gloves, corn husks, and whatever else was in the lagoon. This screen idea didn't work, as it just plugged up with fibrous material. The way to go, in retrospect, was to get a big 3. inlet style Moyno, that would happily deal with whatever was in the lagoon. Given that we are researchers, though, we decided to experiment with a vacuum piston pump.

Me and Guillaume working with his pump Here's Guillaume's invention. We worked together to build a dual-cylinder vacuum manure pump. The idea here was to have the system work like a dual piston pump, but without the headache of piston seals and those associated parts which would have been annihilated by the sand. I built the electronics for it, and helped Guillaume with installation and testing. He wired up the vacuum stuff, and got it actually put together. This was it in the lab, pumping water in a circle, a task which it performed quite well. The principles were sound, I still feel this particular style of pump may actually have some benefit, if the vacuum leak problem could be sufficiently addressed.

The pump on the workbench In my kitchen lab, with the dual-piston vacuum pump set up on the kitchen table. Guillaume was of the opinion that my 555 based timing circuit wasn't giving him correct pulse widths, and I was of the opinion that something else was broken, so out came the scope. It seems overkill to use a 1 GHz scope to measure 8-second style time intervals, but it was the only sampling scope I had in my apartment, where we had met to test. My timing circuit was right, it turns out that there was a vacuum leak, and the pumps suck and blow are not equal. Guillaume thought to fix this by adding a bleeder valve to the piping system

Inside the control box for the pump The internals of the control box. There's three relays, a 555, a power supply, and switch input for manual override and .tank full. switches, which shut off the pump if the fluid level was too high, to keep the pump from ingesting liquids. In typical Linderizing style, the electronics were hand soldered and mounted in a NEMA 4X box, under the belief that we would be installing this thing outside in a rainstorm for the foreseeable future.

The pump pumping away It works! We successfully pumped manure with this pump. However, the dual-cylinder idea did not work quite as well as planned, due to inconsistencies in the manure and bubbles being sucked in the inlet. The timer-based control is not the best way to do this, and the system really needs appropriate fluid-level measurement to reliably pump manure unattended. What ends up happening is one cylinder gradually fills up more than the other, which trips the too-full float switch.

Exploding manure The air compressor Guillaume used was capable of 60 PSI. That white plastic valve behind the wiring conduits was closed when we first turned the pump on. I was sitting in the cleanish spot of the spatter below. Let's see. 60 PSI operating against a closed valve secured with a rubber clamp fitting. Oops.

Face full of not mud Another Big-Baller moment. I'm glad I have a healthy immune system. I was sitting in the path of that pump when it blew off its output coupling, and got hit in the face with the first of many manure-geysers. Shortly after this photo was taken I sozzled my head in a bucket and then went to the barn to hose my face off. This kind of stuff is why Clarkson owes me some new clothes.

The final revision of the manure pump Still going with this vacuum-based pump, and arguing against a number of other unfortunate engineering developments, we turned the flush tank into a larger vacuum vessel. That's Guillaume, with his vacuum pump-in-a-box, in the final implementation of the vacuum-based manure pumping system. This system was working reliably for a while, based on hydrostatic head to load manure from the flush tank into the main digester. Unfortunately, by this time, the tank itself was silted up, and no more pumping was possible.