How to do a percolation test for a soak pit
Dig a small hole where the soak pit will go, soak it overnight and time the drop. That gives the minutes per 25 mm the soak pit calculator asks for.
By Bulan Sarkar · Updated
Bore a hole 100 to 300 mm across, straight-sided, down to the depth where the effluent will soak in. Scratch the sides, put 50 mm of coarse sand or fine gravel in the bottom and fill it with at least 300 mm of water. Come back after 24 hours. If water is still there, set it to 150 mm above the gravel and measure the drop over 30 minutes. If the hole is dry, fill to 150 mm and measure every 30 minutes for 4 hours, refilling as needed, and use the last reading. In sandy soil where 150 mm is gone in under 30 minutes, read every 10 minutes for an hour. Then percolation rate = reading time × 25 ÷ drop in mm, in minutes per 25 mm. A 60 mm drop in 30 minutes is 12.5 minutes. Over 30 minutes the soil won't take a soak pit, and over 60 minutes it won't take a trench either (IS 2470 Part 2, Appendix A and Table 2).
Source: IS 2470 (Part 2):1985 Code of practice for installation of septic tanks, Part 2 Secondary treatment and disposal of septic tank effluent: Appendix A (percolation test), Fig. 1, Table 2, checked Oct 2026.
Before you dig: where, how deep and how many holes
The test tells you how fast your soil takes water at the depth where the effluent will go. IS 2470 asks for the hole to go down to the depth of the proposed trench, or more generally to "any depth at which it is intended to dispose of the effluent" (A-1.1, A-1.2). For a soak pit, that's the part of the pit below the inlet pipe. A hole tested at 300 mm in loose topsoil can give a much faster rate than the clay a metre down.
Dig more than one hole. Clause 3.1 of the code asks for tests "in trial pits at more than one place" so you can take an average. On a house plot, two or three holes spread across the area where the pit could go are enough.
Season matters too. The code warns that the time of year affects the result (cl. 3.2 c), which is why the hole is soaked for a full day first: it brings the soil close to how it behaves in the wettest part of the year (A-1.3). While you dig, note whether water seeps into the hole. If ground water comes within 1.8 m of the surface in the monsoon, Table 1 rules out a soak pit whatever the test says.
The test, step by step
You need a post-hole auger or a crowbar and a narrow spade, a steel tape or a ruler marked in millimetres, a straight stick or a peg as a fixed reference, a bucket of coarse sand or fine gravel, a few buckets of water and a watch.
- Dig the hole: square or round, 100 to 300 mm across, with vertical sides, down to the soaking depth (A-1.2).
- Scratch the bottom and sides with a nail or a knife to break the smeared, polished soil the tool leaves, and lift out the loose crumbs (A-1.2).
- Put about 50 mm of coarse sand or fine gravel on the bottom so the water you pour doesn't scour it (A-1.2).
- Pour water to at least 300 mm above the gravel and leave it for 24 hours (A-1.3). Keep topping it up through the day in fast soil so the sides stay wet.
- Lay a stick across the top or drive a peg beside the hole and measure every water level down from the same point.
- Next day, take the readings. The table below shows which routine to follow.
- Work out the rate from the final reading only: rate = reading time in minutes × 25 ÷ drop in mm (A-1.6).
| What you find | Set the water to | Read | Use |
|---|---|---|---|
| Water still in the hole | 150 mm above the gravel | Once, after 30 minutes | That 30-minute drop (A-1.3) |
| Hole is dry | 150 mm above the gravel, refill when needed | Every 30 minutes for 4 hours | The last 30-minute drop (A-1.4) |
| Sandy soil: the first 150 mm goes in under 30 minutes | 150 mm above the gravel, refill when needed | Every 10 minutes for 1 hour | The last 10-minute drop (A-1.5) |
Why only the last reading counts
When the hole was dry after the soak, the first readings are fast and then slow down as the soil fills up. The code uses only the drop in the final 30-minute period for that reason (A-1.4). The earlier readings tell you whether the soil had settled. If the last two or three are close, it had. If the drop is still falling sharply at the four-hour mark, the soil hadn't settled yet, and the true rate is slower than your last reading.
Write every reading down anyway. The code says the earlier drops can help adjust the method to local conditions, and they're the record you hand to whoever designs the pit.
From the rate to a soak pit, a trench or neither
The percolation rate is the time in minutes the water takes to fall 25 mm (A-1.6). A small number means fast soil. Fig. 1 of the code turns it into the most effluent each square metre of soaking surface can take in a day: Q = 204 ÷ √t litres per m² per day, where t is the rate. Our soak pit calculator uses that formula.
| Percolation rate | Loading, L/m² a day | What the code allows (ground water below 1.8 m) |
|---|---|---|
| 2.5 min | 129.0 | Soak pit or trench |
| 12.5 min | 57.7 | Soak pit or trench |
| 30 min | 37.2 | Soak pit or trench (the limit for a pit) |
| Over 30, up to 60 min | 37.2 down to 26.3 | Dispersion trench only |
| Over 60 min | None | No soil absorption: a filter, sized by an engineer |
If you dug more than one hole, average the rates before you size anything. If one hole is far slower than the others, find out why (a clay lens, a buried rubble patch) before you average it away.
Worked examples
1. Water still in the hole the next morning
- After 24 hours some water is left, so set it to 150 mm above the gravel (A-1.3)
- Drop in 30 minutes: 60 mm
- Rate: 30 × 25 ÷ 60 = 12.5 minutes per 25 mm
- Loading: 204 ÷ √12.5 = 57.7 L per m² a day
- Septic outflow for 5 people at 45 L each: 225 L a day, so 225 ÷ 57.7 = 3.90 m² of soaking wall
- Round pit 1.2 m across: wall 3.77 m round, 3.90 ÷ 3.77 = 1.03 m, plus 0.15 m = 1.18 m below the inlet
Result: 12.5 minutes per 25 mm. A soak pit is allowed; a 1.2 m round pit needs about 1.18 m of depth below the inlet pipe.
2. Hole dry the next morning, clay-loam soil
- Hole is empty, so fill to 150 mm and read every 30 minutes for 4 hours (A-1.4)
- Drops: 38, 30, 26, 24, 22, 21, 20, 20 mm; the last one counts
- Rate: 30 × 25 ÷ 20 = 37.5 minutes per 25 mm
- Over 30 minutes: no soak pit; up to 60 minutes: a trench is allowed (Table 2 Note 3)
- Loading: 204 ÷ √37.5 = 33.3 L per m² a day; all house water for 5 people, 675 L a day ÷ 33.3 = 20.3 m² of trench bottom
- In a 0.6 m wide trench that's 33.8 m, more than the 30 m limit for one run, so two runs of about 17 m, 2 m apart
Result: 37.5 minutes per 25 mm: this soil takes a dispersion trench, not a pit.
3. Sandy soil that drains in minutes
- After the soak, 150 mm of water is gone in 22 minutes, under 30, so read every 10 minutes for an hour (A-1.5)
- Last 10-minute drop: 100 mm
- Rate: 10 × 25 ÷ 100 = 2.5 minutes per 25 mm
- Loading: 204 ÷ √2.5 = 129.0 L per m² a day
- 675 L a day ÷ 129.0 = 5.23 m²; in a 1.2 m round pit, 5.23 ÷ 3.77 + 0.15 = 1.54 m below the inlet
Result: 2.5 minutes per 25 mm: fast soil and a small pit. Keep it at least 18 m from any well, because effluent moves quickly through sand.
Tips and tricks
- Use a scale with millimetre marks and read it at eye level against the same peg every time. On a 20 mm drop, reading 5 mm out either way turns 37.5 minutes into 30 or 50.
- Write the clock time next to every reading. It's easy to lose track of which 30-minute period you're in during a four-hour test.
- Bore the hole rather than chopping it with a spade if you can. A spade smears clay soil, and a smeared hole drains slowly.
- Mark the 150 mm level on a stick held against your peg, so each refill takes seconds and lands at the same height.
- If two holes give very different rates, test a third between them before you average.
- Keep the reading sheet. Whoever designs or approves the pit will ask for it, and so will a plumber if the pit ever overflows.
What we would do
We'd dig two or three holes where the soak pit could go, soak them overnight, and take every reading the code asks for, then put the average rate into the soak pit calculator. If the rate comes out over 30 minutes, or the monsoon water level is within 1.8 m, we'd stop planning a pit and ask a local engineer about a trench or a filter instead.
Frequently asked questions
What is a percolation test for a soak pit?
How do I calculate the percolation rate?
What percolation rate is good for a soak pit?
How deep should the test hole be?
Sources and date checked
- IS 2470 (Part 2):1985, Appendix A (A-1.1 to A-1.6, percolation test), cl. 3.1 a) 4 (test at more than one place, average), cl. 3.2 c) (season), Fig. 1 (Q = 204/√t), Table 1 (ground water within 1.8 m), Table 2 Note 3 (30 and 60 minute limits), cl. 5.2.1, 5.2.2 and 5.3; read from the page images. Checked Oct 2026.
- IS 1172:1993 Code of basic requirements for water supply, drainage and sanitation, cl. 4.1.1 (45 L a person a day for flushing; 135 L for LIG and EWS homes). Checked Oct 2026.
This guide covers measuring how fast soil takes water, for planning. Where the pit or trench may go, how far it must sit from wells and buildings, and any approval belong to your local body and a licensed engineer. Never climb into a pit, septic tank or manhole; the gases inside can kill.