Industries / Construction and property / Commercial and residential construction / Pool construction
Excavation and shell
Excavation and shell is digging the hole and building the watertight vessel that sits in it, whether sprayed concrete, a factory-made fiberglass shell, or a wall kit lined with vinyl.
6
Milestones achieved, from 1934 to 2015.
1
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1934
One pour, no joints
Pools of the 1920s were poured in sections, with expansion joints between the walls and the floor. No sealant yet worked on joints underwater, so these pools "leaked quite a bit", Pool & Spa News writes. Around 1934 Pascal Paddock began pouring the whole shell as one piece. One of the first was a 50 m pool at the Santa Barbara Biltmore Hotel in California, which crews poured over three days and two nights without stopping.
Lessonevery joint in a shell is a path for leaks.
Credit
Pascal Paddock
1940
Gunite pools go volume
Gunite, a name coined in 1912, is concrete blown dry down a hose by the cement gun Carl Akeley showed in 1910, with water added at the nozzle; with it a crew sprays concrete onto steel bars and earth instead of building wooden forms. The American Shotcrete Association cites a Cement Gun Company bulletin on a 1936 pool at the Lehigh Country Club in Pennsylvania, shot against a rock wall, as "one of the first recognizable gunite pools". Pool & Spa News credits Philip Ilsley, who then owned Paddock's company, with the first gunite pool in 1940 in west Los Angeles: a 17 by 34 ft "people's pool" that cost $1,000 to $1,600, one-fifth to one-half the price of his poured pools. His company's output rose from about 40 pools a year to 226. East Coast builders followed later: E.L. Wagner Co. of New York first shot gunite in 1950, Anthony Pools in 1951.
Credit
Philip Ilsley; the Cement Gun Company
1954
Pools sold as kits
Cascade Industries began selling package pools in 1954: concrete-block walls in three sizes, lined with a vinyl sheet, for the colder Northeast and Midwest, where concrete pools cost more to build. Kit walls moved to pressure-treated wood, then to steel and aluminum in the early 1960s; in 1958 Cascade sold a wood-walled kit through the Spiegel catalog for $995. By 1964 package pools made up an estimated 16% of US pool sales.
Lessonconcrete blocks could leave a finished pool as much as 6 in. off the plan; steel walls, cut in a factory, could be made to closer tolerances.
Credit
Cascade Industries; Esther Williams Pools
1950s
Fiberglass shells in one piece
Pascal Paddock is credited with the first fiberglass pool, in the late 1940s: a 15 by 30 ft oval in four parts that fit together like slices of a pie. It sold in small numbers because it was hard to ship and the parts did not always seal or match in color. In the late 1950s W. R. Chance Co. made a one-piece fiberglass pool, 15 by 35 ft. San Juan Pools, founded in 1958, sold molds to dealers for its first two decades, then from the mid-1970s shipped finished shells shaped to stack like paper cups.
Lessonsealing seams made on site was the weak point that one-piece shells removed.
Credit
Pascal Paddock; W. R. Chance Co.; San Juan Pools
2012
A strength floor for shells
In 2012 the American Shotcrete Association's pool committee, chaired by the Connecticut pool builder Bill Drakeley, set a minimum in-place compressive strength of 4,000 psi (27.6 MPa) for sprayed-concrete pool shells, the minimum the ACI 318 building code sets for concrete in contact with water that must have low permeability. It said lower results mean a breakdown in materials or application. Field cases showed how far shells could fall short. In 2005 Steven Gebler and George Seegebrecht of CTLGroup reported cores from a residential pool in the southwestern US that averaged 2,100 psi, 47% of the specified 4,500 psi, with voids 1 to 2 in. below the surface and a water-to-cement ratio near 0.60 instead of the approved 0.45.
Lessontest panels made beforehand can flag bad materials, but the crew's technique on the day sets the shell's strength.
Credit
ASA Pool and Recreational Shotcrete Committee; Gebler and Seegebrecht
2015
Drained pools can float
During a tropical storm in Florida, Nigel Martinez partly emptied his overflowing in-ground pool. The next day the shell had lifted out of the ground, pushed up by water that had gathered beneath it. On 13 February 2015 Florida's Fifth District Court of Appeal held that the water exclusion in his homeowner's policy applied, as courts in New York and South Carolina had held in similar cases. Forensic engineers at Envista list the usual causes: a hydrostatic relief valve (a one-way valve in the floor that lets groundwater in) stuck shut with dirt, and pools built during a drought and emptied after the water table rose.
Lessonthe owner usually carries the cost when a drained pool floats, so the water table matters each time a pool is emptied, not only when it is dug.
Credit
Florida Fifth District Court of Appeal, in Liberty Mutual Fire Insurance Co. v. Martinez
Watch for
A printed pool shell
Contec Australia, a Perth company, says it printed the curved walls of a home pool and spa in concrete in two days, directly from the architects' and engineers' models and without formwork, and calls it Australia's first 3D concrete printed pool. Reports describe the method as print and pour: the printed walls formed both faces, steel bars went between them, and concrete was poured into the gap.
Achieved whensomeone outside the company reports the finished pool holding water through a season, with the shell's strength and cost set beside a sprayed-concrete pool of the same size.
Claimed by
Contec Australia, February to July 2026
Open
01
Where drained pools float
A public table of pools that lifted out of the ground, gathered from court opinions, insurance commentary, news reports and forensic engineers. Each case gives the date, place, shell type, how far the pool was drained, whether a hydrostatic relief valve was fitted and whether it worked, the rain in the seven days before, and the nearest measured groundwater level, so that the conditions that come before a pool floats can be counted.
The briefCredit · unclaimed02
Pool shell strength, tabulated
An open, anonymized table of compressive-strength results from cores drilled out of finished concrete pool shells, with the process (wet or dry mix), age at test, curing, climate, core diameter and core grade, to show how often shells meet the 4,000 psi minimum.
The briefCredit · unclaimed
Open 01 · the brief
Where drained pools float
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public table of pools that lifted out of the ground, gathered from court opinions, insurance commentary, news reports and forensic engineers. Each case gives the date, place, shell type, how far the pool was drained, whether a hydrostatic relief valve was fitted and whether it worked, the rain in the seven days before, and the nearest measured groundwater level, so that the conditions that come before a pool floats can be counted.
Why it is still open
Forensic engineers describe the causes case by case: changes in the land around the pool, relief valves jammed with debris, pools built in a drought and emptied after the water table rose. Courts treat the damage as excluded water damage. A search for a study that counts cases or measures the conditions found none.
Built with what exists
Public court opinion databases; an agent that gathers news reports and cites the page behind each field (see the Note "Findings that cite notes"); NOAA daily rainfall records; USGS groundwater-level records and their equivalents in other countries; a spreadsheet.
First step
Collect the Florida, New York and South Carolina rulings cited in insurance commentary and 20 news reports, and fix the fields from them.
Done looks like
100 cases from 2000 to 2026 in at least three countries, each with at least five of the seven fields filled from cited sources, and a published count of how many followed heavy rain, partial draining or a failed valve.
Open 02 · the brief
Pool shell strength, tabulated
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
An open, anonymized table of compressive-strength results from cores drilled out of finished concrete pool shells, with the process (wet or dry mix), age at test, curing, climate, core diameter and core grade, to show how often shells meet the 4,000 psi minimum.
Why it is still open
The American Shotcrete Association sets 4,000 psi and says cores are commonly expected to reach 85% of the specified strength, 3,400 psi. Published results are single jobs: 2,100 psi in a southwestern pool in 2005, and 1,700 to 2,200 psi in six cores from a homeowner's 2013 pool in ASA's question pages, where a builder had also told a homeowner that 3,000 psi was the "industry standard", which ASA says is not true. A search for a survey that counts how many shells fall short found none.
Built with what exists
Results donated by testing labs, forensic engineers and builders; Shotcrete magazine's open archive on shotcrete.org for published cases; a model such as Claude Fable 5.2 (see the Note "Three working devices") to read scanned lab reports into the table's fields, with a person checking each row; Zenodo to publish it.
First step
Transcribe every pool core result in Shotcrete magazine's open archive and ASA's question pages into the table, each linked to its page.
Done looks like
300 cores from at least 100 pools, each with its source, test standard (ASTM C42 or C1604) and age, and the share below 4,000 psi and below 3,400 psi, broken down by wet or dry mix.