What Is the Difference Between Side and Bottom Water Cooling Design in EV Battery Packs?

Batteries hate heat. Too much of it steals range, ages cells, and raises safety risks. As power density climbs, choosing the right liquid-cooling layout really matters. Today, bottom cooling leads for packaging; side cooling is rising for tighter control. Let’s unpack what actually changes for your pack.

Choose bottom cooling when space is tight and you need proven, modular, cost-effective integration. Choose side cooling when you need stricter temperature uniformity and faster heat response—and your pack has room for added manifolds and piping. Many next-gen packs adopt hybrids (top/bottom or dual-side) to balance both.

I’ll keep it simple and practical—layout first, then thermal results, cost and quick manufacturing notes you can use in real projects.

1. What is side water cooling in EV battery packs?

Think of side cooling as giving the cell two “radiators,” one on each long face. More touch area means more paths for heat to leave. For prismatic cells under fast charge or high C-rate load, that extra contact is a big deal—it flattens temperature peaks and calms hotspots across the module.By the way, the side cooling principle for cylindrical cells is basically the same—just a different shape, same idea, same magic.

In side cooling, plates interleave between cells or clamp to both sides of a module. The larger contact area increases heat flux, delivering lower peak temperatures and more even gradients—ideal for high C-rate discharge or fast-charge windows. However, routing becomes more complex: additional manifolds, pipe bends, and seals raise integration difficulty, packaging volume and potential leak points. It rewards projects prioritising thermal precision over minimal plumbing.

We implemented side cooling in a prismatic-cell battery pack for an electric pickup, clamping cold plates to both long sides of each module. The main challenge wasn’t CFD but manufacturing and assembly consistency—over large contact areas, contact thermal resistance can drift with tolerance stack-up, cell bowing and TIM thickness variation; meanwhile, manifold routing and multiple joints raise integration complexity and sealing risk. Our fixes were pragmatic: use thermal silicone pads to absorb assembly deviations; and design both side plates with zoned, equal-flow manifolds, keeping bends short with generous radii to reduce pressure drop. In operation, the benefits of side cooling’s larger contact area are clear: lower peak temperatures and tighter uniformity—in line with external comparisons on 314Ah prismatic cells showing a clear advantage for side cooling. For pickup and commercial platforms, production-ready side-cooling micro-channel plates are already available and can be tailored to the pack envelope and port layout.

2. What is bottom water cooling in EV battery packs?

Bottom cooling takes the “single radiator” approach: a liquid cold plate sits under the cells or entire module. It keeps the sides free, stacks neatly into trays, and plays nicely with cell-to-pack (CTP/CTC) structures. That simplicity is exactly why it remains the go-to for many prismatic platforms.

By consolidating coolant channels beneath the cells, bottom cooling simplifies the layout and preserves valuable lateral space. It suits modular pack designs and scales cleanly to mass production with vacuum-brazed aluminium plates. While single-face contact can trail dual-side uniformity, bottom plates meet most EV/ESS specs at lower cost and with fewer seals—especially where height/thickness limits dominate.

With fewer seals and a tidy under-cell plate, packaging is straightforward and repeatable. Manufacturers lean on mature processes—stamping or extrusion plus vacuum brazing—to build flat, robust plates that don’t fight the module stack-up. Thermal performance is usually “good enough” for mainstream EV and ESS duty cycles, especially when channel paths are optimised (counter-flow, splitters, or multi-zone layouts). You may not hit the absolute uniformity of dual-side systems, but you often achieve your ΔT targets with less complexity and stronger line yield.

3. How does pack space layout drive the choice?

Start with space, not CFD. If your tray height and thickness are tight, or your width is fixed by crash rails, you already know which way you’re leaning. Side plates need room for manifolds and service paths; bottom plates use the “free” real estate under the cells.

Bottom cooling is generally more compact and easier to integrate where height/thickness are constrained and module interfaces are standardised. Side cooling demands lateral clearance for plates, headers and hoses, plus access for assembly and service. For CTP/CTC layouts chasing high volumetric efficiency, a flat bottom plate often wins; if the enclosure allows extra width and manifold routing, side cooling can be justified by tighter thermal control.

4. How do thermal results compare in practice?

Thermal performance isn’t magic; it’s area plus flow quality. Side cooling brings more area. Bottom cooling catches up with smarter channels. Both can meet tough specs if you design the details and test them honestly.

Studies show dual-side or side-edge cooling reduces Tmax and gradients faster than single-face bottom plates, thanks to increased more uniform heat extraction. Yet, channel innovation narrows the gap: double-layer minichannels, splitters and optimised serpentine paths in bottom plates significantly improve uniformity at modest pressure drop, meeting pack-level ΔT targets for most duty cycles.

If you need the lowest peaks during fast charge, dual-side side plates will usually win—especially with high heat flux near the cell faces. But don’t count out a well-designed bottom plate. Two-layer mini-channels, counter-flow layouts, and split manifolds can pull gradients down impressively without punishing pressure drop. Hybrid concepts (bottom + top or busbar cooling) are also gaining ground, stabilising cells during transient spikes. They work, but they add parts and sealing surfaces—so bake that into your DV/PV plan.

5. Which is more cost-effective to build and scale?

Optimise total system cost—parts, seals, TIM, joining steps, test time, yield—not just the plate price.

Bottom plates leverage mature processes (vacuum brazing, stamping/extrusion cores) and fewer seals, typically delivering lower piece price and better line yield.

● Bottom cooling — cost profile: Fewer components and seals → simpler assembly; mature stamping/extrusion + vacuum/continuous brazing → stable yields; smaller TIM footprint → lower recurring cost; shorter helium/EOL test windows.

● Side cooling — where the cost accumulates: Dual-side plates; extra manifolds/fittings/O-rings; larger TIM or compliant side plates; more welding/brazing and flow-balancing steps → longer leak/EOL testing, lower yield buffers, higher warranty risk control.

● When side cooling still pays: If uniformity prevents derating or extends cell life under fast charge/high C-rates, lifecycle value can outweigh added BOM and test time.

● Cost offsets for side cooling: Make the side plate double as the module sidewall (remove one part); use integrated isoflow manifolds (fewer joints, cleaner distribution); tighten flatness and thin the TIM (reduce recurring cost).

● Decision rule: If packaging is tight and ΔT targets are achievable with smart channels, pick bottom for best $/unit and throughput. If ΔT margins are critical and space allows, pick side (or selective dual-side) and apply the offsets above to control cost.

6. Manufacturing Options for Cooling Plates & Cooling Tubes (Quick, practical)

Cooling Plates

✅Stamped channels + Vacuum/Continuous Brazing (Aluminium).

Ideal for mass production and bottom plates. Few seals, robust flatness, strong yields. Tune land width/depth for the ΔP–ΔT balance; counter-flow helps uniformity.

✅Extruded Micro-channel Plate + FSW.

For higher heat flux or pressure. Dense channels and strong joints (FSW/CMT) make it durable. Great for dual-side modules.

✅Machined Channels + Brazed.

Fast for prototypes; costly for mass production

Cooling Tubes

❌ Extrusion + Bending + High Frequency Welding（HFW)

Although this method exists, it is very difficult to actually produce, because the cooling tube is tightly attached to each battery cell instead of the battery module, which has very high requirements for the bending angle of the cooling tube, and also increases the space requirements in the battery pack

Bottom cooling wins packaging and cost; side cooling wins uniformity—if space allows. Hybrids are rising. Choose by space, ΔT targets, lifecycle cost, and the validation you can stand behind.