# Swimming Pool Loads on a Suspended Slab

> Water is 9.81 kPa per metre of depth, permanent and exact. Depth profiles, tank self-weight, balance tanks, and why an empty pool can govern.

**Category:** Fundamentals  
**Author:** Elena Marchetti (Structural engineer · Founder)  
**Published:** 2026-09-04

---
A swimming pool on a suspended slab is a dead load problem that people file under live load, and the misfiling is what causes trouble. StructLoads carries pool water as what it is: a permanent, precisely known, deep area load that does not move, does not reduce, and does not go away. Water weighs 9.81 kN per cubic metre, so every metre of depth is 9.81 kPa on the slab beneath. A modest 1.4 m deep pool is therefore about 13.7 kPa before you add the tank, the surround, or the plant, which is four to five times a normal office floor.

## The arithmetic that makes pools different

Most floor loads are estimates with margin built in. Pool water is not. The [density of water](https://en.wikipedia.org/wiki/Properties_of_water) is 1000 kg per cubic metre at ordinary temperatures, giving 9.81 kN/m3, and nobody is going to fill the pool deeper than it was built. That precision cuts both ways: the load is certain, so there is no hidden conservatism to lean on, and there is no case in which the pool is empty for design purposes unless emptiness itself is the governing case.

Depth varies across a pool, and this matters more than the average. A [swimming pool](https://en.wikipedia.org/wiki/Swimming_pool) with a 1.0 m shallow end and a 2.0 m deep end applies 9.8 kPa at one end and 19.6 kPa at the other, on the same slab. Averaging to 14.7 kPa across the footprint gives the right total load for the column takedown and the wrong bending moment for the slab, because the deep end concentrates load where the pool floor is also thickest. Treat the depth profile as a varying area load, not a single figure.

The pool structure itself adds substantially. A reinforced concrete tank with 250 mm walls and a 300 mm base contributes its own [self-weight](/blogs/concrete-slab-self-weight-calculation/) at around 24 to 25 kN/m3, which for a 300 mm base is 7.2 to 7.5 kPa before any water. Tiling, screed, and a wet deck surround add another 1.5 to 3 kPa. Add the water and the total permanent load under a deep end passes 28 kPa routinely.

## How to carry a pool through a load takedown

| Method | Best for | Why it works | Main limit | Verdict |
|---|---|---|---|---|
| StructLoads | A pool inside a building takedown | Carries the pool as a varying permanent area load over its true footprint, so the supporting beams and columns pick up the real distribution instead of an average | Does not design the tank, the waterproofing, or the pool shell reinforcement | Best overall for the building side |
| Averaged uniform load over the pool footprint | A quick column total | Fast, and the column axial load comes out close | Wrong slab moments, wrong beam distribution, hides the deep-end concentration | Adequate only as a check |
| Full FE slab model with depth-varying pressure | Transfer slabs and podium decks under pools | Captures the varying load and the surrounding stiffness properly | Setup effort, and the result is only as good as the depth model entered | Right tool for a transfer condition |
| Pool supplier structural sheet | Prefabricated or lined pool systems | States tank weight, fixings, and any point supports the system needs | Stops at the pool, and assumes a rigid support the slab may not provide | Useful input, not a design |

The practical order is to model the pool footprint with its real depth profile, add tank self-weight and finishes as permanent load, decide whether the empty case governs anything, and only then run the takedown to the columns and foundations.

## Why the empty pool can be the governing case

A full pool is heavy and predictable. An empty pool is where two less obvious cases live.

The first is uplift. A pool at basement level with groundwater around it behaves like a boat when drained for maintenance. The same hydrostatic head that pushes on a basement wall pushes upward on an empty tank, and the tank's own weight may not resist it. This is the standard reason pool shells get drained only under controlled conditions with the groundwater relief valve open, and it is a design case, not an operational note.

The second is differential deflection. If the pool sits on a long-span suspended slab, filling and emptying it cycles a large load on and off a structure that is otherwise stable. Long-term deflection under a permanent 28 kPa is considerable, and creep in concrete means the slab keeps deflecting under sustained load. A pool surround detailed to tolerate almost no movement will crack, and the tiling will follow. The load takedown does not solve this, but it is what tells you how large the sustained load is, and therefore how seriously to take the deflection check.

## Plant rooms, balance tanks, and the loads next door

A pool is never just a pool. The balance tank holds a meaningful volume of water and is usually adjacent or below, applying its own hydrostatic load to whatever supports it. Filtration plant, pumps, and heat exchangers concentrate machinery weight into a small room, frequently on the same slab, in the same way that any [rooftop mechanical unit](/blogs/rooftop-mechanical-unit-loads/) does. Chemical storage adds drums of liquid at 10 to 12 kN/m3 in a bunded area that is itself a shallow tank.

The wet deck surround is the load people leave out. A tiled surround on a bedding screed over a fall, plus its waterproofing, runs 1.5 to 2.5 kPa, and it typically covers a footprint two to three times the pool's own area. Add crowd loading on that surround at assembly rates and the deck around the pool can be carrying more total load than a quiet office of the same size.

Movement joints around all of this matter to the takedown in one specific way: they define what is continuous and what is not. A pool tank isolated from the surrounding slab by a movement joint does not share load with it, so the tributary area serving the pool's supporting beams is the tank footprint, not the room. Getting that boundary wrong is the same class of error as mis-drawing any [tributary area](/blogs/tributary-areas-explained/).

## Hydrostatic pressure on the tank walls

The slab carries vertical load, but the tank walls carry a horizontal one, and it follows the same rule as any retained fluid: pressure at depth equals unit weight times depth, acting perpendicular to the wall. At the base of a 2.2 m deep end the wall sees 21.6 kPa of outward pressure, integrating to about 23.8 kN per metre run of wall. The wall spans vertically between the tank base and whatever restrains its top, and if the top is unrestrained it is a cantilever with a base moment near 17 kNm/m.

That thrust does not vanish at the tank base. It arrives as a horizontal reaction into the supporting slab, which has to carry it in-plane to something that can resist it. On a pool sitting inside a building this is rarely a problem because the slab is large and the thrusts on opposite walls balance. On an infinity edge, a freestanding tank, or a pool with a glass wall on one side, the thrusts do not balance and the unbalanced force is real. An infinity edge in particular removes the restraint at exactly the point where the water is deepest.

Emptying reverses nothing, but partial filling during commissioning creates the asymmetric case: one compartment full while an adjacent one is empty, with the dividing wall taking full pressure from one side and nothing from the other. Commissioning sequences are worth writing down for the same reason backfill sequences are.

## Movable floors, diving pits, and the loads they add

A movable floor changes the load model in three ways at once. The mechanism itself is dead load, typically several tonnes of platform and lifting gear, and it applies through point supports at the tank walls rather than spread across the base. Those point loads are permanent and they are eccentric to the wall, so they generate local moments in a wall already carrying hydrostatic pressure.

Second, the water volume does not reduce when the floor rises. Water displaced upward still sits in the tank; it simply moves. The total vertical load on the slab is unchanged, but the distribution over the tank base changes as the floor position changes, which means the base slab has to be checked for the floor at its highest and lowest positions.

Third, a diving pit is a local deep zone. A 4 m pit under an otherwise 2 m pool applies 39.2 kPa over its footprint, double the surrounding pressure, and it is usually located at one end where the structure is already carrying the deep-end concentration. On a suspended slab this is a punching-type concentration, and the same reasoning applies as for any concentrated load compared in [area load versus line load versus point load](/blogs/area-load-vs-line-load-vs-point-load/).

## Fixings, ladders, and the small things that pull

Handrails, ladders, starting blocks, and lane rope anchors all apply concentrated loads to the tank edge, and starting blocks are the significant one. A swimmer's dive applies a horizontal force at the top of a block that is meaningful in local terms and cycles thousands of times. Competition blocks are commonly specified against a horizontal design force and a vertical one acting together, anchored into the tank wall or the deck slab behind it. That anchorage detail is a structural item on a drawing, not a pool fitting, and its load path runs into the same slab edge already carrying the wet deck and the [bulk density](https://en.wikipedia.org/wiki/Bulk_density) of its screed build-up.

## A worked example: a rooftop pool on a transfer slab

Take a 12 m by 5 m pool, 1.2 m shallow end grading to 2.2 m deep end, tank base 300 mm, walls 250 mm, tiled surround 3 m wide on all sides, sitting on a transfer slab over an open plan floor.

Water: the average depth is 1.7 m, so average water pressure is 16.7 kPa over 60 m2, about 1000 kN of water. The deep end applies 21.6 kPa locally, the shallow end 11.8 kPa.

Tank: a 300 mm base at 25 kN/m3 is 7.5 kPa over 60 m2, about 450 kN. Walls at 250 mm and averaging 1.7 m tall around a 34 m perimeter are roughly 14.5 m3 of concrete, about 360 kN, applied as a line load around the tank edge rather than spread over the base.

Surround: a 3 m wide deck around the pool is about 150 m2 at 2 kPa of finishes, 300 kN, plus assembly live load on that same deck.

Permanent total, pool and tank only, is around 1810 kN on a 60 m2 footprint, an average of 30 kPa. On a transfer slab this is the kind of concentrated permanent load that decides the transfer beam depth on its own, and the perimeter wall line load is what governs the local reinforcement. Averaging the whole thing to 30 kPa uniformly across the footprint would understate the edge and overstate the middle, which is exactly the wrong way round.

## The questions to ask before sizing anything

Three questions change the answer materially. What is the true depth profile, including any diving pit or moveable floor? Where is the balance tank and what volume does it hold? And can the pool be drained, when, and with what groundwater relief? A moveable floor system in particular changes the load case set entirely, because it adds mechanism weight, allows a shallow configuration with the same water volume displaced elsewhere, and introduces point supports at the pool wall.

## Key takeaways: pool loads on suspended slabs

Water is 9.81 kPa per metre of depth, permanent, precise, and not reducible. Model the depth profile rather than an average, add tank self-weight and finishes separately, and treat the wall line load around the tank as a line load rather than smearing it. The empty case matters for uplift at basement level and for deflection cycling on long spans. Balance tanks, plant, chemical stores, and a wide wet deck are all part of the same load package, and movement joints define which parts share load.

## Quick answers

A pool is one of the few loads a structural engineer can know exactly. That certainty is worth using properly.

## Sources