How To Choose Storage Racking For A Dark Store
Throughput, SKU density, aisle geometry, and load ratings decide dark-store racking. A practical specification order for operators fitting out urban sites.
A dark store is not a small warehouse. It is a picking machine with a lease attached, and almost every constraint that matters comes from the fact that a picker has to walk to a location, take one or two units, and be gone. Racking that works beautifully in a distribution centre — deep, dense, optimised for pallet movement — will quietly slow a dark store down every hour it operates.
We survey a lot of these sites. The pattern we see most often is a layout borrowed from a conventional warehouse, which maximises stored units per square metre, installed in a building whose entire economic purpose is minutes-to-dispatch. The two goals are not the same, and when they conflict, throughput should win.
Start with throughput, not capacity
The first number worth establishing is not how many SKUs you hold. It is how many orders the site must dispatch in its busiest hour.
That figure drives everything downstream. Peak orders per hour tells you how many pickers are on the floor simultaneously. The number of simultaneous pickers tells you how often two of them need to be in the same aisle. And that frequency is what decides whether your aisles can be narrow and dense, or whether they need to let two people and their trolleys pass without one waiting.
Operators consistently underestimate this. A layout that works with three pickers becomes a queueing problem with six, and the failure is invisible on a drawing — it only appears at peak, when it costs the most. Design for the busy hour, not the average one.
Once you know the aisle geometry your peak demands, you know the floor area left for storage. Capacity is what remains after circulation, not the other way round.
The golden zone is your most valuable real estate
In any picking operation, the shelf heights between roughly knee and shoulder are where a picker works fastest and safest. Everything below requires bending; everything above requires reaching, a step stool, or both.
This band — commonly called the golden zone — should be reserved for your fastest-moving lines. Not your most valuable, not your largest, your fastest-moving. The line picked two hundred times a day belongs at waist height even if it is cheap and bulky. The line picked twice a week can go high or low regardless of what it is worth.
This sounds obvious and is very commonly got wrong, usually because the initial layout is organised by category rather than by velocity. Categories are convenient for replenishment and for staff learning the store. Velocity is what determines picking time. Where the two conflict, a good compromise is category blocks arranged so that the fast lines within each block occupy the golden zone.
The practical implication for racking: you need adjustable shelf heights. A system where levels are fixed at manufacture locks your golden zone to whatever your assortment looked like on day one.
Aisle geometry, and why it is decided early
Aisle width is the single hardest thing to change after installation, because changing it means moving uprights, which means unloading and rebuilding runs. It deserves more thought than it usually gets.
The inputs are the width of your picking trolley, whether two pickers need to pass, whether replenishment happens during picking hours, and whether any mechanical handling enters the aisle at all. A site that hand-picks to a small trolley and replenishes overnight can run genuinely narrow aisles and gain real density. A site where a pallet truck delivers replenishment to the aisle during trading hours cannot, and pretending otherwise creates a bottleneck that no amount of staffing will clear.
Be honest about the second case. Many operators plan for overnight replenishment and then, under volume pressure, end up replenishing continuously. If that is a realistic future for your site, specify for it now.
Load ratings are conditional, and the conditions change
Every load rating is a statement about a specific configuration: a given beam length, a given number of levels, a given upright, load distributed evenly. In a dark store, two of those conditions are unusually likely to drift.
The first is beam length. Relayouts happen often in quick commerce, and widening a bay during one reduces what that bay can carry. The rating that was certified for the original span no longer describes the installation.
The second is distribution. Ratings assume weight spread across the beam pair. Dense single items — liquid cases, for instance — concentrated at mid-span load the beam differently from the same total weight spread across it.
Neither of these is exotic engineering. They just require that whoever moves a beam level knows the rating moved with it. Keep load signage on the racking, keep it matched to the certified configuration, and re-check it after any relayout.
Design for the relayout you have not planned yet
Assortment in quick commerce changes constantly. Lines are added, categories are trialled and dropped, seasonal ranges arrive and leave. The layout that is correct at opening is usually wrong within two quarters, and the store team — not a contractor — is who will change it.
This is the strongest practical argument for fully bolted systems in this application. A bolted rack can be reconfigured with hand tools by the people who work in the store. Shelf heights move, runs extend, and components recovered from a dismantled run fit the new one. There is no service call, no lead time, and no waiting for a slot in someone's schedule.
The trade-off is real: bolted assembly takes longer to install initially than a boltless system, and it requires that bolt tightness becomes part of routine inspection. For a site that will be fitted out once and left alone, that trade often does not pay. For a dark store, it almost always does.
A specification order that works
If you are fitting out a site, this sequence tends to produce fewer regrets than starting from a catalogue:
- Establish peak orders per hour and the simultaneous picker count that implies.
- Set aisle geometry from that picker count, the trolley, and an honest view of when replenishment happens.
- Map velocity across your assortment and assign the golden zone by pick frequency.
- Fix storage type per category — small-bin subdivision for small parts, long-span for cartons, and only then consider anything deeper.
- Confirm load ratings for the configuration you have actually drawn, not the one in the brochure.
- Check reconfigurability — can your own team change this in an afternoon?
Capacity falls out of steps 1 to 4. If the answer at step 6 is no, you will be living with step 3's assumptions for the life of the lease.
Where we usually end up
For most urban dark stores we fit out, the answer is a combination: bolted slotted-angle or long-span shelving for the bulk of the picking face, small-bin subdivision where the SKU count is high and the units are small, and multi-tier structures where the building has height worth using. Pallet racking usually appears only at goods-in, if at all.
The important part is not the product list. It is that the geometry was decided by the throughput the site has to sustain, and that the store team can change it themselves when the assortment moves — because it will.
If you are planning a fit-out and want a second opinion on the layout before you commit, book a site survey. We measure the floor, watch the flow, and tell you what we would specify and why.