Why Concrete Pools Outlast Fiberglass in Clarkston, GA’s Freeze-Thaw Cycle

Fiberglass pool companies spend a remarkable amount of money convincing Georgia homeowners that concrete is old technology. In Clarkston’s specific climate — 30°F diurnal swings in January, 20-plus freeze nights per year, and tree-canopied lots that stay shaded until noon — that marketing is directly wrong on the physics.

The pitch is familiar. Fiberglass installs in three days. Fiberglass is non-porous so it needs fewer chemicals. Fiberglass flexes instead of cracking. Concrete is the horse and buggy; fiberglass is the Tesla. Comparing a well-built fiberglass shell against a bad gunite job on paper, the marketing has a leg to stand on.

Compare them in Clarkston, though — at 1,100 feet of elevation on the DeKalb side of Stone Mountain, where a January afternoon hits 55°F and that same pool deck drops to 25°F by 4 a.m. — and the physics flip. Thermal expansion coefficients matter here. Shell flex ratings matter. The ratio of water absorption between gelcoat and properly sealed plaster matters. “Flexing is better than cracking” is true only if you ignore what flexing repeatedly does to a laminate over two decades.

This post is the honest breakdown. Not a brochure. Not a both-sides-have-merit splitting of the difference. Concrete — specifically properly-engineered gunite — outlasts fiberglass in Clarkston’s climate, and the reasons are measurable in microstrain, psi, and coefficient-of-thermal-expansion numbers that don’t show up in the sales pitch.

Freeze-Thaw Mechanics: What Actually Happens to a Pool Shell at 30°F Below Daytime High

To understand why concrete outperforms fiberglass here, you have to start with what a freeze-thaw cycle actually does to a material. It’s not just ice. It’s repeated dimensional change — thousands of cycles across a shell’s service life — that fatigues one material and barely registers on the other.

Clarkston sits at a 30°F average diurnal swing in January. The afternoon high runs roughly 52–55°F on a sunny day. The nighttime low drops to 22–28°F. On heavily canopied lots — and this city has a lot of them, thanks to the old-growth hardwoods along the Stone Mountain ridge line — the decks and coping stones don’t see direct sun until around 10:30 or 11 a.m. That means the shell material cycles through freezing temperatures longer and more often than an open suburban pool 15 miles west would.

Across a 30-year service life at roughly 22 freeze nights per year, that is 660 full freeze-thaw cycles the pool shell is engineered to survive. Each cycle forces the material to expand as it warms, contract as it cools, and then ride out the stress concentration at any point where it’s mechanically anchored — coping, skimmer throat, return lines, light niches.

Coefficient of Thermal Expansion — The Number That Decides Everything

The coefficient of thermal expansion (CTE) measures how much a material changes dimension per degree of temperature change. It’s the most important single number in this comparison and the one least discussed in fiberglass marketing.

• Fiberglass laminate (polyester resin + chopped-strand mat): approximately 15 to 20 × 10⁻⁶ per °F
• Gunite / shotcrete concrete shell: approximately 5 to 7 × 10⁻⁶ per °F

That is a three-to-one ratio. For the same 30°F swing, a fiberglass shell is moving three times as much as a gunite shell. On a 40-foot-long pool, the difference between those two CTE numbers across a full freeze-thaw cycle translates to about 0.25 inches of dimensional change for fiberglass versus roughly 0.08 inches for gunite. That’s the delta the coping has to absorb, that the bond beam has to restrain, and that the surrounding paver deck has to tolerate without cracking at the joint.

Fiberglass marketing likes to frame this as “the shell flexes.” That’s true. What it doesn’t mention is that the coping and deck can’t flex with it at that rate. The differential movement is what eventually shows up as cracked tile lines, gapped coping mortar, and the telltale hairline separation between the bond beam and the surrounding paver field.

And this is before we get to water. Water is what turns a manageable expansion-and-contraction problem into a structural degradation problem.

Water Absorption, Osmotic Blistering, and Why Gelcoat Eventually Fails

A pool is a waterproof bowl holding 20,000 to 35,000 gallons of water at hydrostatic pressure against its interior surface, 24 hours a day, 365 days a year. For the full service life of the shell. That is an enormous long-duration load that most people underestimate when they hear “fiberglass is non-porous.”

Fiberglass pools aren’t non-porous. They’re low-porosity — which is a different claim dressed up to sound like the same thing. The gelcoat layer, which is what you actually see and touch on the interior of a fiberglass shell, absorbs between 0.2% and 0.4% water by weight across its service life. A properly finished and sealed gunite shell with a premium plaster or pebble interior absorbs less than 0.05%.

The difference sounds trivial until you do the math on what 0.3% absorption means on a 40,000-pound fiberglass shell. That’s 120 pounds of water that has migrated into the laminate over the material’s life. And unlike a concrete shell — which is engineered to be slightly alkaline and slightly porous on purpose, in equilibrium with the water chemistry around it — fiberglass resin is hydrolytically reactive. Water doesn’t just sit in it. Water reacts with it.

The technical name for what happens is osmotic blistering. Water migrates through the gelcoat, reaches the polyester resin underneath, and reacts with uncured styrene and other soluble compounds in the laminate. That reaction generates dissolved acids and glycols in micro-pockets inside the shell wall. Those pockets exert osmotic pressure, which pulls in more water, which accelerates the reaction. Eventually — typically at the 12- to 18-year mark in warm climates, sooner in cold-cycle climates like Clarkston — those pockets deform into surface blisters. Sometimes small and chalky. Sometimes dime-sized and soft. Occasionally the size of a quarter, filled with yellow-brown vinegar-smelling fluid.

This is not a defect. This is what the material does over time. Better manufacturers use vinylester or epoxy barrier coats behind the gelcoat to slow the rate. Cheaper brands use standard polyester and call the eventual blistering a “warranty exclusion related to water chemistry.” Both statements are technically accurate.

What Water Does to Concrete Instead

A properly mixed gunite shell — 4,000 psi compressive strength at 28 days, water-to-cement ratio of 0.45, air-entrained to around 5–7% for freeze-thaw resistance — isn’t reactive with water the way fiberglass is. It’s designed to be in equilibrium with it.

When water migrates into the concrete matrix, it carries calcium hydroxide and other alkaline compounds with it. Those compounds reinforce the cement paste over time rather than degrade it. This is called autogenous healing, and it’s one of the least-appreciated properties of engineered concrete: minor micro-cracks, on the order of 0.004 inches or less, actually seal themselves through continued hydration of unreacted cement particles. A fiberglass laminate has no equivalent mechanism. Once the resin matrix is compromised, it stays compromised.

Shell Flex, Coping Stress, and the Truth About “Fiberglass Flexes — It Doesn’t Crack”

The most repeated fiberglass marketing line in North America is that fiberglass flexes, and because it flexes, it won’t crack during a freeze. The claim has a kernel of truth. A fiberglass laminate is more ductile than a rigid gunite shell in short-term dynamic loading. Drop a hammer on each of them and the fiberglass will bounce; the gunite will chip. Fine.

The problem is that freeze-thaw isn’t short-term dynamic loading. It’s long-duration, low-amplitude, cyclical fatigue. And fiberglass hates cyclical fatigue. Every material engineer working in composites knows this. Polyester laminates have a fatigue endurance limit that’s approximately 25–30% of their ultimate tensile strength, compared to 50%+ for steel-reinforced concrete. Meaning: fiberglass loses load capacity faster than concrete does when you stress it repeatedly at sub-ultimate loads.

In short-term impact, fiberglass wins. In long-term freeze-thaw, concrete wins. And freeze-thaw is what actually decides a pool shell’s service life in Clarkston.

What this looks like on an aging Clarkston fiberglass pool, typically at year 14 or 15: the gelcoat begins to develop microfractures at high-stress zones. Usually you see them first around the skimmer throat, around the light niche, and along the radius of the step bench. These aren’t dramatic cracks. They’re hairline spider networks, visible only in raking morning light. They get worse every winter. By year 20, the gelcoat is chalking, the laminate is delaminating in one or two localized areas, and the homeowner is being quoted $8,000–$14,000 for a gelcoat restoration that will buy another 6–8 years before the same fractures return.

A well-built gunite shell at year 20 needs a plaster refinish. That’s it. Not a structural repair — a cosmetic refinish. $6,000 to $9,000 in Metro Atlanta at current pricing, and the new plaster rides on top of a shell that is structurally identical to the day it was shot. The concrete doesn’t fatigue. It just gets slightly more carbonated on the surface, which is addressed by the refinish process anyway.

Coping Stress — The Part Nobody Warns You About

Coping is the transition piece between the shell and the deck. Travertine, bluestone, bullnose pavers, poured concrete cantilever — whatever the surface, the coping is what absorbs the differential movement between the shell and the deck during temperature cycling.

On a gunite pool, the coping moves a fraction of a millimeter per cycle. The mortar bed underneath it is engineered for that tiny movement. On a fiberglass pool, the shell is trying to move 0.2 to 0.4 inches per 100 linear feet, but the coping bonded to it is rigid. Something has to give. Over 660 freeze-thaw cycles in a 30-year span, what gives is the bond between the coping and the shell. You see it first as a hairline gap in the caulk line. Then as a loose coping stone. Then as a full coping replacement job at year 22, which is not cheap.

This is why fiberglass pool manufacturers almost universally recommend flexible cantilever-edge concrete decks around their shells — not bonded coping stones. They know what coping stress does to their product. The sales rep just doesn’t bring it up unless you specifically ask.

Field Evidence: What Clarkston Pool Service Lives Actually Look Like

Physics is one thing. Job-site evidence is another. Here’s what we see when we’re called out to service aging pools in Clarkston neighborhoods — the Montreal Park district, the older lots along Indian Creek, the shaded parcels up near Stone Mountain Park.

Fiberglass pools installed in the 2004–2010 boom years are now hitting their 15- to 20-year service window. Shells installed by good crews on well-drained lots with upgraded barrier-coat gelcoats are mostly still serviceable — interior cosmetic work and coping re-bedding, but shells intact. The ones installed on tight shaded lots, backfilled with native clay instead of engineered fill, and finished with standard polyester gelcoat are the ones we’re being called about. Blistering on the waterline tile band. Hairline spider cracks at the light niche. Coping lifted off the bond beam by 1/8 of an inch. One Clarkston client had a fiberglass shell pop upward an inch one wet January — not a shell failure but a hydrostatic pressure relief failure. The shell had no dead weight to resist buoyancy when the water table came up.

Gunite shells have the opposite problem profile. The issues we see at the 20-year mark are almost always surface-finish issues — plaster chalking, tile grout failure, a broken fiber-optic light lens — not structural ones. The shell itself sits there, dead-weight anchored against hydrostatic uplift by its own 35,000-plus pounds of reinforced concrete, and the steel cage in the walls still reads full continuity on a megger test.

Access and Delivery — The Clarkston-Specific Issue

There’s a practical argument for fiberglass that deserves its own mention because it’s the one time fiberglass’s case is strongest: access.

A fiberglass pool arrives on a flatbed truck as a one-piece 40-foot monolith. It needs a 12-foot-wide crane access corridor from the street to the excavation. Some pool models require 14 feet. On Clarkston’s older interior streets — Montreal Road, Brockett Road, the cul-de-sacs off Market — that access corridor often doesn’t exist. The lot itself may be fine. The path from the street to the lot isn’t. Power lines. Canopy trees planted in 1978. Fences and fence gates sized for a lawn mower.

A gunite pool doesn’t have that limitation. Gunite arrives through a boom-pumped hose running from a truck staged on the street, and the hose needs 3 feet of clearance. Steel rebar is carried in by hand. Forms are built on site. Nothing oversized ever has to move through the yard. For roughly 40% of Clarkston’s older infill lots, that’s the difference between “we can build you a pool” and “you’ll need to remove three trees and pull the fence on both sides of your neighbor’s yard.”

So fiberglass marketing is correct that it installs faster on paper. It’s quietly incorrect that it installs at all on many Clarkston lots. Access decides the project before materials ever do.

The Recommendation — Why We Only Build Concrete in This Climate

Primetime Pools GA builds gunite pools. Not because we’re ideologically opposed to fiberglass — the product has its place, mostly in sandy coastal soils with mild winters and wide-open lot access — but because when we run the engineering against this specific climate, this specific soil type, and this specific access profile, gunite wins on every axis that matters over a 30-year horizon.

Here’s the honest summary. If you live in Clarkston, on a typical DeKalb County lot with tree canopy, red Piedmont soil, and the standard winter cycle of freezing nights followed by 55°F afternoons, the concrete pool you have built in 2026 should still be structurally sound in 2056 with one mid-life plaster refinish. That refinish is budgeted and expected. The shell itself is a 30- to 50-year asset.

The same lot built with fiberglass will likely need a gelcoat restoration at year 14 to 18, a full coping re-bed at year 20 to 22, and a meaningful decision at year 25 about whether to spend another $12,000 on interior restoration or start over. That’s not a failure of the fiberglass manufacturer. It’s the material doing what composite laminates do in freeze-thaw climates. The math is the math.

The other factor — one we haven’t hit yet — is design freedom. Gunite is formed in place. Any geometry is buildable: integrated tanning ledges with fiber-optic bubblers, perimeter overflow edges, sharp-radius geometric shapes, a spa spilling into the pool at a specific horizontal elevation, raised bond-beam walls with sheer-descent water features. Fiberglass is a molded monocoque. You pick from the manufacturer’s catalog of 40 or 50 stock shapes, and any deviation adds a seam and a warranty exclusion. For a Clarkston client who wants a pool that looks like it belongs to their house — not a pool that looks like every fiberglass pool in DeKalb County — that matters almost as much as the engineering does.

None of this means fiberglass is a bad product universally. It means it’s the wrong product for this climate, this soil, and the design expectations most Clarkston homeowners bring to a custom pool project. Pick your tool for the job. In this specific job, concrete is the right tool.