Nine thousand years of basement drainage, in one article

The question the system answers.

When the ground around a foundation gets saturated, three things happen, and they have always happened. Oxygen leaves the soil and the roots above it suffocate. Hydrostatic pressure builds against any wall in the way. And the water, having no easy path out, finds the cheapest one available, which on a modern house is usually a cove joint, a mortar gap in a block wall, or a hairline crack in concrete.

The people who first noticed this were farmers, not waterproofers. They lost crops to it. Some of them lost civilizations to it. What follows is the long story of how the answer got worked out, and the short story of why the answer is buried around your foundation right now.

The ancient world figures out drainage.

Agricultural drainage shows up in the archaeological record around the ninth millennium BC. The earliest version is simple, and not far off from what an honest contractor would still call a French drain: a trench dug into wet ground, partially backfilled with the most permeable material the builders could find. Gravel, stones, bundles of brush, straw. Anything that left voids for water to move through. The ground above stayed dry enough to plant. The ground below stayed wet enough to be someone else's problem.

By roughly 4000 BC, Mesopotamian farmers were cutting open ditches to route ponded runoff. By 2500 BC, the same was happening in the Indus Valley. Around 2000 BC, the Indus Valley engineers fired the oldest drainpipes we have ever found, hand-formed clay tubes laid end-to-end in trenches under what passed for streets. They look like drainpipes because that is what they are. The geometry barely changed for the next four thousand years.

The Mesopotamians who skipped this step paid a heavy price. They irrigated heavily without draining, the water table rose, mineral salts wicked to the surface, and the fields salinized. There is a serious historical argument that the slow agricultural collapse of southern Mesopotamia is in large part a drainage failure. Wherever the irrigated water has no path out of the soil profile, the soil eventually goes too saline to farm.

A few thousand miles east, the Liangzhu culture in the Yangtze River Delta was solving the same problem at terrifying scale. Around 5,100 years ago they moved millions of cubic meters of clay to build a 30-kilometer network of dams, levees, and reservoirs. They did it without draft animals; the material was hauled by watercraft. The system controlled monsoon flooding, irrigated their rice paddies in the dry months, and kept the urban settlement above water year-round. It is one of the largest engineered hydraulic systems of the ancient world, and almost nobody outside of Chinese archaeology has heard of it.

The Romans get more credit than they probably deserve because their drainage writing survived. Cato the Elder around 200 BC and Pliny the Elder in the 1st century AD both describe a technique that should sound familiar to any modern installer. Dig a trench. Lay curved clay roof tiles in it, arch-up, so the tile holds the trench cavity open. Backfill with gravel. The tile is the conduit, the gravel is the filter, the trench slope moves the water. Two thousand years later we have replaced the clay tile with perforated PVC and the unsorted gravel with washed 3/4-inch stone in a geotextile sock, but the structural logic is unchanged.

The Romans figured out that the trench cavity, the open conduit, and the surrounding filter all do different jobs, and that all three jobs have to be done. Nothing about that has changed.

The Romans also showed what the same engineering can do in a city. The Cloaca Maxima, built around 600 BC under the Etruscan kings, drained the marshy Velabrum valley by backfilling the swamp with 20,000 cubic meters of soil and gravel and routing the local streams through masonry-lined conduits into the Tiber. By the 1st century AD, the overflow from Rome's eleven aqueducts was sent through the same system as a continuous flushing mechanism. The Roman Forum sits on what used to be wetland. That is a foundation drainage project, just at city scale.

Elkington's crowbar and Smith's gridiron.

The story jumps eighteen hundred years, and not much happens. Medieval European farmers used the same techniques the Romans had used. Curved clay roof tiles were salvaged from old buildings and laid in trenches, again and again, for centuries. The methods stayed empirical. Nothing was written down systematically.

That changes in the late 1700s in Britain. Population is rising fast, food demand is rising faster, and reclaiming bogs and waterlogged clay soils for cultivation is suddenly worth real money. Two very different farmers solve very different versions of the problem, and between them they produce most of the engineering that still defines our trade.

Joseph Elkington, the crowbar moment

Joseph Elkington was a Warwickshire farmer who in 1764 was digging an interceptor trench across the bottom of a hillside to stop spring water from flooding his pasture. He had figured out the basic mechanism: rain landed on porous high ground, sank vertically through sand or gravel, hit an impermeable clay layer somewhere below, and then traveled sideways under hydrostatic pressure until it found the lowest opening in the slope. That opening became the spring. The spring became the bog. The bog killed the pasture.

His standard four-foot interceptor trench was not deep enough to catch the source. The water was clearly coming from somewhere lower than he could safely dig. So he picked up an iron crowbar, drove it down into the bottom of the trench, and punched through the clay layer beneath. Water came up.

                |   |   |
                v   v   v   (rainfall)
          #################### (porous sand/gravel hilltop)
          #   ............   #
          #   . groundwater. #
          #   ............   #
   =======#==================#========= (ground surface)
          #                  #   |
          #                  #   |  (lateral hydrostatic flow)
          ####################   v
          #::::::::::::::::::#
          #::  interceptor  ::#
          #::    trench     ::#  <-- excavated drain
          #::::::: || ::::::::#
          #:::::::|||::::::::#
          #:::::::|||::::::::#  <-- vertical auger borehole
          #:::::::|||::::::::#      taps water-bearing stratum below
          #.......|v|........#

That moment, the bar through the clay, is the same engineering principle as a weep hole drilled through a basement block wall to release water trapped inside the block voids. Same problem. Same answer. Find the trapped water, give it somewhere to go.

Elkington spent the next thirty years refining it. He swapped the crowbar for a ground-boring auger, learned to read the local geology before he excavated, and drained enough farmland to attract national attention. In 1795, worried that his failing health (he had epilepsy) would prevent him from passing his methods on, the British Parliament voted him £1,000 and a gold ring on the condition that he share what he knew. The Edinburgh land surveyor John Johnstone was commissioned to write it down, and in 1797 published the first systematic technical account of subsurface drainage in English. Late in his career Elkington partnered with Lancelot “Capability” Brown to drain ornamental landscapes on aristocratic estates, which is how the technique entered polite English consciousness.

James Smith of Deanston, the gridiron

Elkington's method had a built-in limit. It required a hillside, a known spring, and a clay layer to puncture. On flat clay plains, where the waterlogging came from slow downward infiltration rather than lateral spring pressure, it did nothing.

James Smith of Deanston was a Scottish cotton mill manager who took over his uncle's farm in 1823 and immediately ran into exactly that problem. Smith's answer, which he started publishing in 1831 and which the British agricultural crisis of 1834 turned into industry standard, was to stop hunting for hidden springs and treat the whole field as the problem.

He laid parallel drains 15 to 30 feet apart, 2.5 to 3 feet deep, across the entire field. Every field. Every time. Standard spacing, standard depth, standard install. To get the rain into the drains, he invented a heavy subsoil plow that shattered the impermeable clay layer beneath the topsoil without bringing it to the surface. Water now had a vertical path down and a horizontal grid to carry it away.

The basement perimeter drain we install today is Smith's gridiron rotated ninety degrees. The wall is the field. The footing is the row spacing. The dimple membrane is the subsoil plow. Same logic. Same hydraulics. Different orientation.

Smith's system won the long argument with Elkington's because it was systematic and repeatable. Elkington needed a skilled geologist on site for every job. Smith's gridiron worked on any waterlogged ground a contractor could measure. By the 1850s, parallel-drain layouts were the global default. They still are.

The actual French in 'French drain'.

Here is the thing almost nobody knows. The French drain is not French. It is named after a New Hampshire lawyer.

Henry Flagg French was born in Chester, New Hampshire, in 1813. He practiced law in Exeter and Boston, sat as a Justice of the Massachusetts Court of Common Pleas from 1855 to 1859, and later served as Second Assistant Secretary of the U.S. Treasury under Ulysses S. Grant. By all biographical accounts a serious legal mind. Also a farmer with a deep technical interest in agricultural science and an unusual amount of patience for tile drainage.

In 1857 he took an agricultural tour of Europe, traveling through England, Scotland, Ireland, Belgium, and Prussia. He compared the leading drainage practices of the era: Smith's Deanston gridiron, the Keythorpe system (drains following soil-undulation lines), the Wharncliffe system (interceptors at slope bases), and several variations on tile-and-trench. He took notes, measured, asked questions, and went home.

In 1859 he published Farm Drainage: The Principles, Processes, and Effects of Draining Land with Stones, Wood, Plows, and Open Ditches, and Especially with Tiles. The book was a synthesis: not original engineering as much as careful integration of the best European practice for an American audience. But it contained one move that was his own, and it is the move that matters.

The graduated aggregate filter

French had observed that most subsurface drains failed for the same reason. Fine soil particles migrated into the gravel backfill over time, plugged the voids, and choked the drain. The conduit was fine. The pipe was fine. The void space around the pipe got slowly clogged with silt and the whole system lost capacity.

His solution was to sort the gravel by size on the way in. The coarsest stone went directly against the central clay-tile conduit, where high flow velocity could carry away any small particles that did reach it. Progressively finer gravel went outside the coarse core, with the finest material at the boundary where the gravel met the native soil. The grain-size step-down acted as a natural filter, catching fines at the outer ring and never letting them reach the conduit.

            ================================== (ground level)
                          |
                          | (infiltration)
                          v
            ##################################
            #   fine sand/gravel exterior    #
            #   ==========================   #
            #  |  coarse gravel core      |  # <-- graduated aggregate filter
            #  |   _________________      |  #
            #  |  /                 \     |  #
            #  | |   roofing tile    |    |  # <-- curved terracotta tile
            #  | |     conduit       |    |  #     laid with 1/8-inch gaps
            #  |  \_________________/     |  #
            #  |   ||             ||      |  #
            #   ===||=============||======   #
            #      v             v           # <-- water inflow via gaps
            ##################################

This is the move. Every modern drain tile install does the same thing, just with better materials. The clay conduit is now perforated PVC or corrugated HDPE. The graduated gravel filter is now a non-woven geotextile filter sock wrapped directly around the pipe, plus a uniform bed of washed 3/4-inch stone around the sock. The geotextile does the fine-particle filtration that French was doing with sorted sand. The clean washed stone gives the water its high-permeability path to the pipe. The function is identical. The materials are better.

The other thing French codified was slope. A drain has to move water fast enough to carry sediment away but not so fast that the discharge erodes. He recommended a hydraulic gradient of about 0.5 to 1 percent (a vertical drop of 1 foot per 100 to 200 feet of run), calculated by the now-classic relation:

where S is the slope, h_f is the vertical drop, and L is the run. Every modern interior drain tile install slopes its perimeter pipe to that same range, toward a sump pit. The math is one hundred and sixty-five years old. The math still works. Some of the original drains French installed in New England in the 1850s have been reported to still be functioning over a century later.

Scotch Johnston's Folly.

The same year French was finishing his book in Massachusetts, an immigrant Scottish farmer in Seneca County, New York, was already twenty years into proving the point empirically.

John Johnston was born in 1791, came to America from Scotland, and in the 1830s bought a heavy-clay farm in Fayette he named “Viewfields.” The soil was the kind of cold, slow-draining ground that defines basement waterproofing trouble two centuries later. Planting was always late. Frost-heaving destroyed crops in the marginal seasons. Johnston remembered seeing fired clay drain tiles in Scotland and reasoned the same trick would work on his farm.

In 1835 he ordered two pattern tiles from Scotland and took them to Benjamin F. Whartenby, a crockery maker in Waterloo, New York. Whartenby fired three thousand tiles. Johnston laid them across his land in 1838.

            =============================== (ground level)
                        |     |
                        v     v (water percolation)
            ###############################
            #                             #
            #    excavated clay soil      #
            #    backfill                 #
            #                             #
            #      ________________       #
            #     /  cylindrical   \      #
            #    |   clay tile      |     #  <-- fired clay pipe
            #    |   conduit        |     #      laid butt-to-butt
            #     \________________/      #
            #                             #
            ###############################

His neighbors thought he was an idiot. They called the project “Scotch Johnston's Folly,” predicted the tiles would freeze and shatter under the New York winter, predicted they would clog with mud, predicted he had ruined his soil for good. Johnston ignored them.

His first crop after the tile install came in at fifty bushels of wheat per acre. His undrained land had been giving him five. By 1852 he had installed sixty miles of clay tile under three hundred acres. By the time he retired, the figure was seventy-two miles. His neighbors stopped laughing.

Johnston is the reason every American farmer in the second half of the nineteenth century knew what drain tile was. He is also, by a direct lineage, the reason every basement in the Twin Cities has the kind of perforated pipe it does. The product changed. The idea did not.

By 1848, John Delafield had imported the first mechanical tile-making machine from England. By 1871 there were ten tile factories in Waterloo, New York alone. John Dixon introduced horse-drawn tile machines in 1851. Concrete drain tile came in 1862. The clay-and-concrete tile industry boomed for the next hundred years.

Materials drifted with what local farmers could get. A Swedish immigrant named John Olson in Swedesburg, Iowa, built his early drains from hollowed-out walnut logs when clay tiles were not yet available in his region. As soon as commercial clay tiles reached Iowa, he ripped out the wooden lines and replaced them with the proper fired tubes. The improvisation tells you how strong the demand was. People used whatever was on hand because the drained land paid for itself in a single season.

The 1967 pivot to plastic.

For a hundred and twenty years after Johnston, drain tile was made of clay or concrete and laid by hand. Then a small group of researchers and an extruder operator in Delaware ended the clay-tile industry in about ten years.

In 1955, Dr. Glenn O. Schwab at The Ohio State University published Plastic Tubing for Subsurface Drainage, the first serious academic case for replacing fired clay with low-density polyethylene. The argument was straightforward. Plastic was lighter, cheaper to transport, did not rot, did not break in cold weather, and could be made in long continuous lengths instead of short brittle sections.

In 1961 German manufacturers invented continuous corrugation extruders, the machines that produce thin-walled plastic pipes with ring stiffness high enough to survive being buried. In 1965, Drs. James L. Fouss and Norman R. Fausey at the USDA Agricultural Research Service ran field trials on the German corrugated pipe and confirmed it survived agricultural soil chemistry as well as clay did. By 1967, a company called Advanced Drainage Systems started commercial production of corrugated polyethylene pipe in Middletown, Delaware. They picked a 4-inch diameter for the simple reason that the U.S. drain-tile industry was standardized on a 4-inch clay tile, and they wanted farmers to be able to swap one for the other without rethinking the install.

Corrugated HDPE weighed one twenty-fifth what clay tile weighed. A laser-guided drain plow towed behind a tractor could install twenty thousand feet of it in a day, with no manual trenching at all. By 1978, the clay-tile industry that John Johnston had midwifed in 1838 was essentially over. The new pipe was everywhere.

The 4-inch perforated pipe wrapped in geotextile sock that we install around your foundation right now is the residential descendant of that 1967 agricultural product. The diameter is the same. The corrugation pattern is the same. The way it sits in a bed of washed stone, sloped to a sump, is identical to how it sits in a 1967 cornfield. The basement is a smaller, vertical version of the same problem.

Lifting the water out, the sump pump.

All of the above assumes you can move water downhill to somewhere it can leave under gravity. Farms can do this; you ditch the field and run the discharge to a creek. Foundations below the natural water table cannot. The gravity drain collects the water at the lowest point in the basement and stops. Something has to lift it from there to outside the building.

The mechanical solution arrived in pieces over about a hundred and fifty years.

                       discharge outlet (to daylight)
                              ^
                              |
                    __________|__________
                   |          |          |
                   |     [check valve]   |  <-- prevents backflow
                   |          ^          |
                   |          |          |
                   |     [   pipe   ]    |
   ================|==========|==========|=============== [concrete floor]
                   |          |          |
                   |    .-----|-----.    |
                   |    | [pedestal] |   |  <-- pedestal motor (old-style)
                   |    |  [ motor ] |   |
                   |    '-----|-----'    |
                   |          |          |
                   |   .------|------.   |
                   |   |  [centrifugal] |  <-- submerged volute assembly
                   |   |  [ impeller  ] |
                   |   '----------------'
                   |     |||||||||      |
                   '---------------------'
                       (sump basin)

Before 1859, people bailed. Wooden hand pumps and cast-iron lever pumps were sunk into hand-dug basement sump pits, and the work was done by whichever family member drew the short straw. It worked the way you would expect. Slowly, with effort, badly when you needed it most.

In 1859, James B. Lansing built the first steam-driven dewatering pumps for industrial mines. The same year, LaBour Pumps introduced its heavy-duty vertical steam sump pump for factories. The steam pumps worked, but they were not residential. They were too big, too expensive, and they needed a coal-fired boiler running alongside them. A homeowner could not have one.

In 1875, the patent attorney Edward H. Knight recorded the term “sump pump” in his American Mechanical Dictionary. This is when the category exists as a named idea in English.

In 1886, the Penberthy Injector Company of Detroit, Michigan introduced the “automatic cellar drainer.” This one is delightful. It used no electricity. It used the Venturi effect. A copper float in the sump pit, as it rose, opened a valve that let pressurized municipal tap water flow through a small nozzle in the discharge line. The high-velocity jet created a vacuum at right angles, which sucked the cellar water into the stream and pushed both out of the building. As long as you had municipal water pressure, you had a sump pump that did not care about a power outage. Penberthy water-powered backup pumps using the same principle are still sold today, and they still work for the same reason.

In 1928, Charles E. Weir patented the first electric automatic centrifugal sump pump. The motor sat on a pedestal above the water line, with a long vertical drive shaft running down to a submerged impeller. The pedestal sump pumps you still see in older Twin Cities basements are direct descendants of Weir's design.

In 1950, Harry Goodman, co-founder of the Little Giant Vaporizer Company, patented the first fully submersible water pump. Goodman's innovation was a hermetically sealed, oil-filled motor housing that could run completely underwater. No drive shaft. No pedestal. Quieter, more compact, less electrically dangerous. Andrew Shane filed a competing residential submersible design at almost the same time. The submersible took over residential basements through the postwar housing boom and has been the standard ever since.

Every sump pump in every Twin Cities basement we install is some version of Goodman's 1950 submersible, paired with a battery backup that solves the one weakness of the design (it stops working when the grid goes down). The water-powered backup pumps we occasionally see in older homes are still Penberthy's 1886 idea. The drive shaft on the rare old pedestal still in service is Weir's 1928 design. There has been very little real invention in sump pumps since the war. There has been a lot of refinement.

Why your basement drain looks the way it does.

Walk into a properly waterproofed Twin Cities basement and what you see, if you can see anything, is a thin black strip of dimple membrane against the wall, a sealed lid on the sump basin in the corner, and a discharge pipe heading out through the rim joist. Underneath the floor, where you cannot see it, is the part that matters: a 4-inch perforated pipe in a bed of washed stone, wrapped in a geotextile filter sock, sloped a half a percent toward the basin. The submersible pump in the basin lifts the collected water through a check valve and discharges it to daylight or to storm sewer, depending on what your municipality requires.

                         |      |
                         v      v (surface infiltration)
              ##############################
              #   backfilled soil          #
              #   ======================   #
              #  | washed gravel bed    |  #
              #  |  __________________  |  #
              #  | /  perforated PVC  \ |  # <-- perimeter drain
              #  | |   (French drain)  | | #
              #  |  \_________________/  |  #
              #   ======================   #
   ===========#============================#======== (footing level)
              #                            #
              ##############################
              [     foundation footing     ]

Now look at where every piece came from.

• The trench and the gravel bed are nine thousand years old. Mesopotamian farmers were digging the same trench with the same backfill in 7000 BC.
• The conduit holding the trench open is the Roman curved roof tile, replaced after 1859 by clay drain tile, replaced after 1967 by perforated polyethylene pipe. Same job, three materials, two thousand years.
• The graduated filterbetween the conduit and the soil, the geotextile sock plus clean stone, is Henry Flagg French's 1859 graduated aggregate idea, executed with better materials.
• The half-a-percent slopeto the sump is French's hydraulic gradient calculation, unchanged since he published it.
• The perimeter layoutaround the foundation footing is Smith of Deanston's 1831 gridiron, rotated for vertical geometry.
• The submersible pump and float switchare Harry Goodman's 1950 patent, with battery backup added to cover the power outage Goodman couldn't.
• The water-powered backup pumpif you have one is Penberthy's 1886 Venturi-effect cellar drainer.
• The dimple membrane on the wallis the modern equivalent of Smith's subsoil plow, opening a vertical path so water at the wall can drain harmlessly down to the perimeter pipe instead of seeping into the basement.

That is the underpinning of the entire foundation-waterproofing industry. Almost nothing in it is new. Almost everything in it has been continuously refined for thousands of years. When we quote a homeowner an interior drain tile install, we are quoting the synthesis of about nine thousand years of accumulated engineering, expressed in modern materials.

Knowing the lineage matters for one practical reason. If a contractor is selling you something that does not connect to this lineage, that does not look like the engineering above, you should ask why. A baseboard channel that sits on top of the footing rather than at footing depth is not what nine thousand years of trial and error converged on. A no-slope drain is not what French wrote down. A non-sealed sump basin is not what Goodman intended. A sump discharge that empties into your sanitary sewer is not what any modern municipality permits. These shortcuts come up constantly in residential quoting, and recognizing them is the easiest way to tell whether the system you are buying is doing what the engineering says it should.

The work is old. The materials are new. The principles, the trench, the conduit, the graduated filter, the slope, the lift, are exactly the same. We are still doing what the Romans were doing, what French wrote down, and what Goodman patented. We just have better pumps.

If you ever stand over your sump pit in March and watch the pump cycle on every two minutes during the spring thaw, that is nine thousand years of engineering doing its job. It is older than writing. It is older than the wheel. It is, by a fairly wide margin, the most continuously refined piece of practical engineering humans have ever produced. It is also probably the single most important system in your house, and you cannot see most of it.

That is the trade. That is what we do for a living.