In brief
A pool on the lowest floor can be built safely even where groundwater sits high — provided the decision is made before the house is designed, not after. Three things have to be calculated first: the highest seasonal water level on the plot, the uplift acting on the buried structure, and the moisture balance of the pool hall. The underground part is then usually built as a white tank — reinforced concrete designed to hold the water back by itself. Most of the work sits in the concrete mix and in the details; a coating applied at the end will not do the job on its own. Adding a pool to a finished design can sometimes be done, but it ends in reworking the structure, the floor heights and the services — costs that did not have to happen.
“The water is too high”
“We had two contractors already, both walked away — they say the water is too high.” I have heard that sentence more than once, and I understand both sides. A contractor does not want to risk something he has not built before — which is honest of him, as far as it goes. And you start to suspect you bought a plot with a hidden defect.
Defect is too strong a word. Groundwater is one of the plot’s parameters, the same way soil bearing capacity or frost depth are. You measure it, you design for it, and the house has to respect it. The trouble starts when someone designs as if the water were not there — or declares “it can be done” or “it can’t” before anyone has calculated anything.
High water is more common than people think: at the foot of slopes, near streams, on clay ground that will not let water soak away. The table can sit just below the lawn and climb higher still in spring. If a pool is to go on the lowest floor of such a plot, the basin and the rooms around it will spend their whole life below the water level. That part of the house behaves a bit like a hull — water pressing on it from the sides and from below, all the time. It can be calculated and it can be built. You just have to start calculating before the design.
Three things to calculate before the design exists
First: the highest water level on the plot
The first thing is a proper picture of what is happening underground. Drilling a borehole “because the permit needs one” is not enough here. You need surveys done for this specific purpose: soil layers, permeability, the water level and how much it moves.
That movement is the heart of it. A reading taken in a dry August can show the water far lower than where it will stand in March after the thaw. You design for the highest level — which is why the water is observed across more than one season, and on a serious basement the surveyors often leave piezometers on the plot (a standpipe in the ground where you can see how the water sits — nothing spectacular, but a year of those readings tells you more about a plot than many an expert opinion). I will not tell you how much water your plot will have in March — nobody honestly can, until it is measured. Where long-term data is missing, a design water level is set with a margin, because the structure cannot be redesigned after the fact.
There is one more thing few people think about: the water level on a plot is not fixed for life. I once worked for clients who had bought an old house to renovate. The basement had been bone dry for years, so nobody gave it a thought. A few years after the renovation, the municipality rebuilt the road. The works damaged the field drains that carried water away from the surrounding land, and their outflow ended more or less at the line of that plot. The water rose until the basement stood flooded almost halfway up. Nobody on that plot had neglected anything — the conditions around it simply changed. That is why, with a pool below ground, I design with a margin and in such a way that the structure defends itself. I will come to that in a moment.
There is a formal side too. Under the Polish rules on foundations, a water table above foundation level puts the building into what the regulations call complex ground conditions. The house stops being “a regular building with a basement” — the geotechnical category is decided by the designer together with the geotechnical engineer, but for a building like this it means in practice a wider set of studies: alongside the geotechnical opinion, a documented site investigation and a geotechnical design. From where you stand, that is not bad news. More things get calculated on paper before anything stands on site.
Second: uplift
A structure sitting below the water table displaces water, and the water pushes back on the slab from underneath — all the time, not just in spring. And here is the part that surprises almost everyone: the dangerous case is not a pool full of water. Try pushing an empty, capped bottle underwater in a bathtub and you will feel it — a full one sinks, an empty one shoves your hand back up. A pool basin behaves much the same. Full, it helps, because its weight holds the structure down. It gets dangerous empty — drained for maintenance, say, right when the water outside is high. The other bad moment is the build itself: the tank already stands, but the floors and walls that are supposed to weigh it down are not there yet. The structural engineer has to have both answers calculated in advance — the mass of the structure, ballast or anchoring, and the sequence of works when the water is high.
One thing I say out loud, because it tends to be passed over: drainage is not the answer to uplift. Drainage lowers the water level for as long as it works. Systems fail, pipes silt up, power can be gone for two days. A design in which the only thing protecting the structure is a working pump is, to my mind, irresponsible — a pump is not structure. I treat drainage as support and relief, nothing more.
Third: the moisture the pool itself produces
The third thing surprises people the most, because it is not about groundwater at all — it is about the water in the basin. The surface of warm water gives off moisture without pause, and that moisture has to have somewhere to go. The moisture balance of the pool hall is calculated together with the structure, at design stage, because the results drive the duct cross-sections, the equipment and the floor height. More on this below.
Let me also say straight away who should be doing all this arithmetic, because on this subject it is easy to assume “someone will surely calculate it”:
- water level and its swings — the geologist or geotechnical engineer;
- geotechnical category and the scope of studies — the designer with the geotechnical engineer;
- uplift — the structural engineer;
- watertightness of the underground part — the structural engineer with a concrete technologist;
- moisture balance and hall ventilation — a pool ventilation designer; that is a specialisation of its own, not “the installer will handle it while he’s at it”;
- the plant room — the architect with the pool technologist.
If any of these roles has no name against it in your team, that is exactly where something will give later.
The white tank — concrete that holds the water itself
An underground structure that lives in water is not sealed “with something on the outside” at the end of the build. It is usually designed as a white tank: reinforced concrete that is the barrier — the same concrete carries the house and holds back the water. The name may sound exotic, but the idea is plain: the watertightness has to live in the structure itself, and painted-on layers are an extra at most.
For that to work, three things have to come together.
The mix and the cross-sections. Concrete with a low water-to-cement ratio, with controlled heat of hydration, with a watertightness class chosen with a margin for permanent water pressure. And slabs and walls thick enough — a thick section is spare distance for the water: it presses into the concrete from outside, but it is meant to stall deep inside the wall, far from the side you live on. The thickness is not a number from a catalogue; it comes out of calculations for the actual water pressure.
Crack control. Reinforced concrete cracks by nature — you cannot switch that off, you can only govern the size of the cracks. So it is designed so that crack widths stay within the value set in the design, typically up to 0.2 millimetres. Only cracks that fine have a real chance of sealing themselves. This is governed by anti-shrinkage reinforcement and by curing the fresh concrete. That second part sounds trivial, but a fresh tank that nobody remembers for its first week can undo the engineer’s whole calculation.
The details. White tanks almost never leak through the middle of a wall. They leak at the joints: construction joints, service penetrations, the places nobody planned. So construction joints are planned to the metre and sealed systematically — waterstops, coated steel plates, swellable profiles — before they exist. And penetrations through the tank are kept to as few as possible, each with its own collar.
When the tank is poured, when the joint seals are inspected and when the basin is tested for tightness, I am on site in person. Not because I do not trust people. Because these are stages that cannot be fixed afterwards — concrete does not wait.
And protection on the outside? A well-designed white tank formally does not need it, and there are builds done without it. I am more cautious than that. I choose the protection of the underground part case by case, and I set it up in layers — several technologies working together, not one. If one fails, the house keeps defending itself. A plan B and a plan C, of sorts. Because buried work is something I never want to come back to — least of all when coming back means deep excavation around a finished home.
The pool hall is a climate of its own
I promised to come back to the moisture, so here it is. You know this from your own bathroom: after a long shower, the mist settles on the mirror and the window — the coldest surfaces in the room. A pool hall does exactly the same thing, only without pause and on a larger scale, because the surface of warm water evaporates day and night. If that moisture is not led into the ventilation in a controlled way, it will find the coldest spot in the structure and condense there. For years, quietly.
So the pool hall gets its own ventilation and dehumidification system, independent of the rest of the house. A few of its parameters are not a matter of taste:
- relative humidity is held at around 50–60 percent, with 64 treated in design practice as the hard upper limit;
- the air is kept two to four degrees warmer than the water — warmer air draws less moisture off the surface, and whoever climbs out of the pool does not feel cold;
- supply air is run along the glazing and the cold walls, not straight over the water — an air stream across the surface speeds up evaporation, which is exactly what we are fighting;
- the exhaust works on the opposite side of the hall, and the whole room is kept at a slight negative pressure against the living areas, so that moisture and the smell of pool chemistry do not wander through the house.
On top of that come the things renderings simply do not show: the vapour barrier and the envelope details designed for this one microclimate, and equipment built for a damp, corrosive pool environment — a higher corrosion-resistance class, sealed electrics. Standard ventilation kit can corrode here within a few years.
And one practical question I ask surprisingly early: will the basin be covered when nobody is swimming? A closed cover visibly cuts evaporation — it changes the moisture balance, the sizing of the unit and the running costs of the whole system. Which is why that question comes up at design stage with me, not after the first season.
The balance itself is not something I calculate personally — that is the pool ventilation designer’s job. My part looks different: making sure he gets real data, that the design has room for his ducts and his plant room, and that nobody “optimises” his assumptions along the way. The climate itself — the dew point at the glazing, dehumidification and energy — I take apart separately, in the piece on the climate of an indoor pool.
The plant room and the things you do not see
Water treatment, pumps, filters, the ventilation unit with dehumidification — all of it has to live somewhere. And here three things surface that are easy to forget on a floor plan. Volume with a margin, because the equipment will one day be replaced and that replacement cannot mean breaking walls. A separate service route, so the technician does not walk through the living quarters. And acoustics calculated against the rooms where people sleep — pumps and ventilation run long hours, nights included (and a bedroom can land directly above the plant room; you cannot hear that on a floor plan).
You can tell a well-designed pool by the fact that the house does not know it is there. No pumps to be heard anywhere, no smell of chlorine, dry glass.
Where this subject usually goes wrong
With pools on underground floors, the same mistakes come around again and again. It is almost always the order of decisions:
- Surveys from a single season. The design was done “dry”, the water arrived in spring. Everything after that is firefighting.
- A pool drawn into a finished design. Structure, floor height, ventilation, plant room — every one of them has to be reworked. Designing the pool in from the first line costs a fraction of that.
- Ventilation “like a house, only bigger”. Residential heat-recovery ventilation does not dehumidify and is not built for pool chemistry. The mistake shows first on the glass, later in the walls.
- A floor that is too low. The hall’s ducts need space under the ceiling. If nobody calculated that before the height of the underground floor was fixed, you end up with a room that holds the pool — and no room above it.
- Drainage as the only protection. As above: a pump is not structure.
I am not writing this to scare anyone. Every one of these scenarios comes from the same thing: somebody started designing before the surveys. Reverse the order and they disappear.
When an underground pool is not worth it
There are cases where, after the surveys, I am the one advising a change of plan. I would rather have that conversation before the design, because after the design it costs far more.
After proper surveys I almost never say “it can’t be done.” If the pool matters to you, the question is what protecting the building will cost — not whether I can build it. What I do say plainly is when a similar result can be had more sensibly: moving the pool to ground level, or into a separate, shallower building. I lay out both routes with their consequences, and the decision stays with you. How that arithmetic turns into a price that does not change during construction is covered in my text on the lump sum.
A separate case: the house design already exists, and the pool is to be “added” to it. A pool on an underground floor is not a room you bolt on. If the structure, the heights and the services were not calculated for it, the honest answer is: not in this design. Redesign first, pool second.
There are also a few things I do not promise. A pool on high water will not cost the same as one on low water. It will not be maintenance-free either — none of them are. And you will not hear “it can be done” from me before the ground and groundwater surveys, because it would be guesswork.
What I ask before I say yes
If you are considering a basement pool on a plot with high water, these questions will come up anyway. Better that they come up before the design:
- Are there ground and groundwater surveys from more than one season, and do they show the highest water level we must design for?
- Is the decision “pool: yes or no” being made before the structural design of the house?
- Who calculates uplift — and are the empty-basin and construction-stage scenarios calculated?
- Is the underground part designed as a white tank, with a plan for construction joints, service penetrations and tightness tests?
- Has the hall’s moisture balance been calculated by a pool ventilation designer — and does the design have room for the ducts and a plant room with its own service route?
- Are the structural engineer, the pool technologist and the ventilation designer working on one model of the building before the permit design is drawn?
- Who will physically stand there when the tank is poured and when it is tested for tightness?
This is the list I run through myself on every such project. If any of these questions has no answer, the design is not ready — however good it looks in the renderings. And questions about working with me — from budget to the pricing model — are answered in the Questions section.
Frequently asked questions
Is a basement pool even possible on a high water table?
Yes. High water does not rule out a basement pool — it changes how the design is done and raises the cost of the underground part. After proper surveys the question is usually “how much”, not “whether”. One condition: a structure calculated for the water from the start.
Is drainage enough to protect against groundwater?
As the only protection — no. Drainage works for as long as it works: a pump failure or silted pipes bring the full water pressure straight back onto the structure. I treat it as support for a structure that defends itself against the water.
What surveys are needed before designing a pool on high water?
A ground and groundwater investigation done for this specific purpose: soil layers, permeability, the water level and how it moves with the seasons. Observation across more than one season; on a deep basement, piezometers on the plot. Only on that data do the design decisions get made.
When is it better to choose something other than a basement pool?
When the cost of protecting the underground part stops making sense against the alternatives: a pool at ground level, or in a separate, shallower building. That is arithmetic I show after the surveys — with the consequences of both routes, and the decision stays on your side.
This material is educational and describes typical solutions — it does not replace surveys or design work. Every project requires its own ground and groundwater investigation, detailed design and supervision.
If you are planning a pool in difficult ground conditions, bring the subject to me before the architect starts drawing — while it is still possible to choose between solutions instead of rescuing one. Let’s talk about your project.