Category Archives: Well

Most essential if you want to dig a well: gravel pump design

The key tool for sinking a well (unless you want to climb in and dig it out) is a device called a plunscher, or a gravel pump. A punscher is a simple metal pipe, with a rubber flap (valve) at the bottom, and it will fill with sand when pulling it up quickly. This has to be done with enough speed to suck in sand, and you can add water to the well to get this done quickly (plunscher only works when submerged – at least mostly – in water). While this works well for sand and small gravel, the more efficient tool is a gravel pump – essentially a plunscher with a piston inside that will actively suck-in the sand from the bottom of the pipe. The piston will lower again under its own weight, and by repeated pulling it gravel pump will fill with sand and gravel quickly.

This is the setup, you can see the blue rope, make sure to use a good and strong rope, because the forces involved are quite substantial and sand will wear down these ropes, so better to exchange them from time to time in case you want to dig multiples wells.

First modification, a washer mounted in the middle of the rubber flap, it ensures better tightness and valve action. Some plunschers have this feature already from the supplier. Better to use some 4 mm fibre-reinforced NBR rubber.

The mounting screws are difficult to reach, so I mad an extension from a piece of wood…

The weld quality of the plunscher wasn’t all that good, but well, this is not a rocket engine, but a tool fabricated to a certain (rather moderate price tag). This price tag is also the reason why you wouldn’t directly by a gravel pump: a good gravel pump will set you back 150 EUR, whereas a plunscher can be obtained for 45~50 EUR. And, surely, it is more fun to do some modification yourself rather than buying all the expensive tools right away.

This is the modified plunscher, you can see the piston, and the cut-out.

The piston needs to have a valve action as to allow the piston to sink at moderate speed, but obtaining good vacuum when pulling the rope. I also tried to use two ropes: one for the piston, one for gravel pump case, but these ropes get entangled and there is no need for such ropes: the weight of the pump will keep it down, if you just pull with the right force and speed.

The piston needs to maintain some clearance from the wall, otherwise, it will get stuck with small stones, etc.

As a seal, I used leather from an old school bag, very firm and thick leather, and a somewhat smaller rubber disc (fairly hard NBR rubber). This worked well with no significant wear. The leather can be made such that there is almost no gap, for example, 1 mm, to ensure strong suction for fine sand. If you have coarse sand, probably you can also work with a larger gap or worn piston seal, you just need to pull the rope more often. I generally recommend to keep the gap small unless you run into some trouble with specific gravel or some particular sand or stone.

At the bottom, for the last meter, I attached a serrated, rather dangerous-looking teeth ring, to cut into the ground and loosen the sand. It worked marvelously. Generally I can recommend to keep the well flooded as much as possible by adding water all the time, then no sand will flow into the pipe if you hit a somewhat “liquid” layer. I probably kept the water level at least 50 cm above the water table.

The rope needs to be mounted such that the piston cannot be pulled out completely, this needs to be adjusted properly. Sure you could also weld a guide ring or similar for the piston, but it worked out very well with the more “wobbly” piston, the vacuum is strong, and the extraction of sand was more limited by the nature of the lowest sand and clay layer, rather than by the vacuum level.

It seems it wouldn’t hurt for the gravel pump to be a bit heavier, for example, by using a longer and heavy-walled tube, but this will also require more force to lift it up. Also the piston could be a bit heaver to sink more quickly, but well, you will figure out how to operate with the tool after a little while, and a shorter gravel pump is certainly more easily handled. Just make sure to wear proper work boots, because your toe may crack if you drop the gravel pump on it. An the serrated front will bite into your foot as nicely as it bites into hard sand.

As for the diameter, it seem that the 89 mm outer diameter is well suited for a DN115 well pipe. I could imagine that with a larger gravel pump, it may be difficult to withdraw, and will be overly heavy. So unless you gain other experience, any pipe around 90 mm outer diameter will work. I would suggest to use 88.9×5.6 which is a bit heavier, rather than the 88.9×3.6 used for the punscher. But probably all will work if you handle it right.

Down from the well: all kinds of sand

Most important for a well is the nature for the water-bearing layer. Already when using the soil drill I noticed coarse sand when hitting the water layer, and it is so liquid that it can be easily washed down and has almost no turbidity. This is basically a good finding, intermediately coarse sand, with little fines.

The top layer had a few larger stone, but in the water bearing layer, the largest were maybe about 8 mm in diameter.

Interestingly enough, the was a solid but -fortunately- thin layer of sand solidified by white matter in about 3 meter depth (just about 1 cm thick layer). Maybe the river running here over my land dried up some 1000s of year ago?

To color of the sand is somewhat red, but this color doesn’t wash out. Diffing further, from about 6 meters down, gray to black sand appeared. This sand was considerably finer and pretty difficult to remove by plunsching, so I used plenty of water and many strokes of a gravel pump to remove it (I modified the plunscher to a gravel pump by adding a piston – will be described in another post).

Finally, at about 7.5 m down, the sand turned more and more black with some brownish clay fragments and plenty of mica (shiny particles).

Given the 0.3 mm slot width of the filter, about 1-2 mm sand would be quite ideal as a water bearing layer, so I was definitely happy to hit solid clay about 7.6 meters down, and sunk the lowest pipe some 10 cm into this clay layer. Note that the clay layer is really heavy pliable clay, it doesn’t seem to swell or dissolve in water easily. So I even decided not to further close the bottom of the well pipe, it seems soundly stuck and closed by the natural clay.

Some study of the sand revealed that the 7 meter sand has quite some sharp and irregular particles that can clog the filter, so better to keep the filter out of this area as much as possible. Maybe the lowest filter section (2 m in total) is now in the black sand layer for about 0.5 meters.

7 meter sand:

Even more important in such case to not overload the filter, to keep sand from getting into the well by keeping the inlet velocity at well below 0.03 m/s, better 0.02 m/s, which is possible by taking about 2000 L/h through the 2 meters of filter section.

The sand in the main water bearing layer is looking much better, it is coarser, and has more rounded shape.

Red marked are some small gravel, and the red lines show millimeter distances.

Main water bearing sand at about 5 meters, microscopic picture:

For thoroughly removing all sand from the well, I used a sand sucker construction from regular PE pipe (32 mm outer diameter), and a 8 mm pneumatic hose inserted such that it is pointing upwards, and extending about 15 cm inside of the pipe. With ample supply of pressurized air (from a compressor) connected, it will pump up a mixture of water and sand even to 8 meters, no problem. Sure it is a mess of water, sand and dirt, but it is an easy way to get rid of all the fine sand and mica that can’t be effectively removed by the gravel pump finally.

I also supplied plenty of water by a flushing tool, finally, also used this to soften the clay and to flush out a few cm of clay, but introducing a fairly high pressure (6 bars) water jet and pumping out the dirty water at the same time.

I continued to pump out water with the air-operated pump for about 1 hour, finally, I connected an old electric pump to the well for about 3 hour, and during all that time there was basically clean water after the first 30 minutes. And so far it doesn’t show any sand residues after one week of use, so maybe we are safe. Let’s check in one year! Surely I will keep all the tools so I can flush out any sand or residues in coming years should need be.

Drilling and digging: a new well

Since I have moved to my new house, there are extensive gardens that need plenty of watering these days. So far, I have been using a 1984 driven (abyssinian) well, merely a 1-1/4″ steel pipe rammed into the ground. This well has been struggling to provide enough water, less than 500 L per hour. Probably it has reached the end of its life and all attempts to rejuvenate it helped for a while, but I have been looking for a more permanent solution.

That’s the old well – now closed with a cap!

Rather than building again a driven well, which is not quite suitable for the quantities of water that I am looking for, say, 2000-3000 L/h rate, I decided to try a drilled (open) well. Size DN115, which is 125 mm outer diameter, 5 mm wall pipe, specially designed for wells (GWE well pipe, PVC-U, K-series).

The first 80 cm were easily dug with a spade, it is mostly sandy soil with some gravel stones.

Further, I needed to use a soil drill. A neighbor provided it generously, and with some old pipes extensions were made to reach to about 6 meters.

The soil here in the Rhine valley is quite suitable for these kind of drills, in just two hours or so, and with plenty of sweat, the whole reached down to the water table at around 4 meters. Still removing some sand, but you can’t drill into a mixture of water and sand… it will just form a cavity.

There are other kinds of drill, but this is a close-up, a large corkscrew.

Before we proceed, we need to insert the pipe, now assembled to 4.8 meters: 0.8 m sump, 2 meter filter pipe (0.3 mm inlets), 2 meters of plain pipe. It is not all that heavy and I managed to get it in quite easily.

Now, we can already see almost to the center of the earth, at least, 4 m closer to the center…

The pipe needs to be securely mounted so that it can’t move around too much.

Next, we have to deepen the well by a process called plunsching. Bit by bit removing the sand from the inside, and lifting it up with the device, which is basically a steel pipe with a valve at the bottom.

It worked well with coarse sand, but with finer sand, I needed to tighten up the seal and modified it a bit to close tightly. Otherwise the fine sand tends to run out.

Also needed to make a special tool to reach to the screw at the bottom. All a bit inconvenient, but it works.

Also critical is the loading of the pipe, first, I added about 150 kg, later about 240 kg. Easily managed by some old concrete pavers that are about 10 kg each.

The plunscher, I attached it with 3 chain links to the rope, this held it better in place and it could be handled easily.

With up to 350 kg, (the load an my own body weight from time to time), we are well in the save area of the weakest link, the filter pipe.

According to the manufacturer, 2 meters of the filter pipe used should be good enough for nearly 4000 L/h, I may take 2000~2500 L/h, so there is a good safety margin

It is critical to stay below about 0.03 m/s water inlet speed, otherwise there may be effects detrimental to the lifetime of the well.

The cuts in the filter pipe are pretty precise, hard to do this at home.

After about 6.5 meters, things got really difficult, with fine sand, which was also pretty much solidified. I used various tools including water hoses and a mud sucker (a pipe with a PU hose inserted, pointing upwards inside the pipe). The mud sucker uses compressed air (I just supplied the full amount my compressor can generate) and at the top a mixture of sand and water will come up. It is a little mess, but convenient to operate. Also I added plenty of water to the well to keep the level as high as possible, otherwise further sand may be sucked in.

Finally, I reached a layer of clay, and with the help of large quantities of water and air, I managed to dig some 20 cm into it, but it seems really solid and pliable clay.

This scheme shows the well as sunk. it is about half-filled with water, and the suction point is located between the two filters, in an area of no inlets. Ideally, the inlet should be above the filter pipe, but I wanted to allow at least 2 m of water column above the inlet, and with the pump outside the well, the turbulences and local load on the filter pipe will be minimal.

The inlet is just a section of pipe, with many holes drilled into it.

The distribution system and piping as mostly done with 32 mm cold-water PE 100 pipe, connected to legacy 3/4″ piping of my workshop and garden, and some newer pipe (16×2 Pipetec composite pipe).

After only just a few minutes with an old pump to remove dirty water, already the water became nice and clear. Maybe because of the thorough work with the sand pump, there was not much dirt to remove. Also I decided against closing the bottom of the pipe, which now seems to be very solidly embedded in clay anyway.

The water is pouring out plentifully, it is pleasure to the eye and a delight forever!

Everything else could be done easily, just mounting a few pipes and machining a lid from 30 mm PVC plate.

Finally, protected it with some concrete plate and stones while providing easy access for removing water from the pipework in winter, basically, just opening the check valve at the top.