Racking layouts: selective vs drive-in vs VNA, and the space/throughput trade-off

A warehouse quoted at AED 45 per sq ft in Jafza and one quoted at the same rate in Al Quoz can hold entirely different numbers of pallets, and the racking layout is why. Selective racking gives roughly one pallet position per 25-30 sq ft of floor space once aisles are counted in; drive-in racking gets that down to 14-20 sq ft per position (Hammerhead, Selective vs. Drive-In Racking, retrieved 2026-09-06). On a 20,000 sq ft shed at Dubai South rents, that gap is worth several hundred pallet positions you are either paying for and using, or paying for and leaving empty as aisle.

Run your own footprint through the warehouse space calculator before reading further — the number that comes out for a given pallet count changes by 30-50% depending on which of the three layouts below you assume, and the rest of this article explains why.

Key Takeaways

• Selective racking needs 8-13 ft aisles and gives 100% pick access to every pallet; it is the default for anything with more than a handful of SKUs.
• Drive-in racking needs no internal aisles and packs 40-60% more pallets into the same footprint, but only the front pallet in each lane is reachable, so it runs LIFO.
• VNA (very narrow aisle) racking uses 5-6 ft aisles worked by turret trucks and gains 30-50% more positions than selective, while keeping full pallet-by-pallet access.
• VNA's density comes at the cost of specialised trucks, wire or rail guidance, and a taller building — it is a capital decision, not just a layout one.
• The right layout is set by SKU count and pick frequency, not by rent per sq ft alone; the wrong layout on cheap land still costs you in retrieval time.

Selective racking is the default because it asks the least of you

Selective racking is a single-deep, single-pallet-position rack served from a standard aisle. Every pallet faces an aisle directly, so every pallet is reachable without moving anything else. That is the entire case for it: 100% selectivity, compatible with any FIFO, LIFO or lot-tracking discipline, and it runs on whatever forklift you already have.

The cost is aisle width. A sit-down counterbalance truck needs roughly 12-13 ft to turn into a rack; a reach truck brings that down to 8-10 ft (Hammerhead, Narrow-Aisle Selective Pallet Racking, retrieved 2026-09-06). Aisles are the single biggest use of floor space in a selective layout, and they hold no stock at all. For a business with a broad SKU range, frequent picking, and mixed pallet heights, that trade is still the right one — the flexibility is worth more than the square footage it costs.

Drive-in racking trades access for density, and the trade is a real one

Drive-in racking removes the aisle between rows and lets pallets nest several deep into the rack structure itself, loaded and retrieved by a forklift driving directly into the lane. Total capacity is lanes × depth × levels — eight lanes, five pallets deep, four levels comes to 160 positions in a block that would hold far fewer under selective racking (NTL Storage, Drive-In Racking Pallet Capacity, retrieved 2026-09-06).

The constraint is access order. Standard drive-in loads and unloads from one end only, which forces LIFO: the last pallet in is the first one out, and everything in front of a given pallet has to be cleared before you can reach it. Drive-through racking opens both ends and allows FIFO, but that means aisles on both sides of the block, which claws back much of the density gain — a five-deep drive-through block runs at roughly 1.8-2.0 pallets per square metre of footprint against roughly 2.4 for the equivalent drive-in (NTL Storage, retrieved 2026-09-06). In practice, most drive-in installations settle at 80-85% effective utilisation rather than the theoretical maximum, because LIFO forces some positions to sit empty when the active SKU does not fill a lane, and a working buffer of 10-15% is normal rather than a sign something has gone wrong.

Drive-in earns its keep on high-volume, low-SKU-count stock with a long shelf life — building materials, bagged goods, anything bought and sold in full pallet lots. It is the wrong choice the moment a lane needs to hold more than one SKU or one batch date, because the pallets behind the front one are, for operational purposes, unavailable.

VNA racking is selective racking with the aisle shrunk to the truck, not removed

Very narrow aisle racking keeps the single-deep, fully selective layout of standard racking and simply narrows the aisle to what a specialised truck needs rather than what a standard forklift needs. A turret truck rotates its forks 90 degrees inside the mast instead of turning the whole vehicle, which lets working aisles run as narrow as 5-6 ft — collapsing a 12 ft counterbalance aisle to a 6 ft turret-truck aisle frees roughly 600 sq ft per 100 ft of rack run (Hammerhead, Narrow-Aisle Selective Pallet Racking, retrieved 2026-09-06). Because every position still faces an aisle, VNA keeps 100% selectivity — you are not giving up pick access to get the space back, unlike drive-in.

The reported density gain varies by source and by how tall the building goes: Hammerhead puts it at 30-50% more positions than a standard selective layout, while Krost Shelving cites up to 50% more storage capacity than standard racking systems (Krost Shelving, VNA vs. Selective Pallet Racking, retrieved 2026-09-06). Either way, the gain compounds with height: turret trucks routinely lift to 30-40+ ft, well beyond what a reach truck handles comfortably, so a VNA layout is usually also a taller building.

That height and that aisle tolerance are where the bill arrives. True VNA aisles at 5-6 ft leave the truck inches of clearance either side, which means wire-in-floor or rail guidance is close to mandatory, not optional. The floor itself needs to be flatter than a standard slab, because a taller mast magnifies every dip the truck rolls over. And turret trucks cost more to buy and to maintain than reach trucks, need certified operators, and are usually specified with a backup unit or a service contract, because a single truck failure stops every pick in the aisle it serves (Krost Shelving, retrieved 2026-09-06). None of that shows up in the racking quote — it shows up in the equipment and civil works line, which is why VNA is a capital-planning decision as much as a warehouse-design one.

The footprint difference is the number that actually moves your rent bill

VNA (very narrow aisle) racking is a storage layout that keeps single-deep, fully selective pallet access, like standard selective racking, while narrowing the working aisle to what a turret truck needs rather than what a standard forklift needs.

<!-- [ORIGINAL DATA: worked footprint and annual rent comparison across selective, drive-in and VNA layouts for a fixed 2,000-pallet-position warehouse] -— Take a warehouse that needs to hold 2,000 pallet positions. Run the same pallet count through selective, drive-in and VNA assumptions and the required floor area moves by a wide margin before you have added a single square foot of office, dock or racking cost — the aisle width and the density factor are doing all the work. At AED 40-45 per sq ft, which is the prime rate quoted for Jafza North and South through H2 2025 ([Knight Frank, The UAE Industrial and Logistics Report H2 2025](https://www.knightfrank.ae/newsroom/article/2026/2/the-uae-industrial-and-logistics-report-h2-2025), retrieved 2026-09-06), a footprint difference of even 10,000 sq ft between layouts is worth AED 400,000-450,000 a year in rent alone, before touching the racking or the trucks. That is the number worth putting through the [warehouse space calculator](/calculators/warehouse-space-calculator) with your own pallet count, your own SKU mix and your own rent quote, rather than assuming selective racking is the safe default because it is the familiar one. Where rent is closer to Al Quoz's AED 85-100 per sq ft for Grade A stock ([Knight Frank](https://www.knightfrank.ae/newsroom/article/2026/2/the-uae-industrial-and-logistics-report-h2-2025), retrieved 2026-09-06), the same footprint saving is worth roughly double, which is usually the point at which the VNA equipment bill stops looking expensive and starts looking like the cheaper option over a five-year lease. ## Matching the layout to the operation, not the other way round The decision runs on two questions, not one: how many distinct SKUs or batches does a single lane need to hold, and how often does the far pallet in that lane need to move before the front one does. A single-SKU, slow-turning bulk product answers both questions in favour of drive-in. A broad SKU range with mixed batch dates answers both in favour of selective or VNA, and the choice between those two then comes down to whether the rent saved from the tighter aisle is worth more than the turret trucks and the taller building it requires. This is also where the logistics setup decision connects to the wider operational build — racking layout, dock configuration and inbound flow get specified together, which is why it belongs in the [logistics and shipping setup](/income/accelerator/launch/logistics-shipping) planning rather than being bolted on after the lease is signed and the shell is already built to one ceiling height. Mixed layouts inside a single building are common and often correct: drive-in for the bulk, slow-moving lines at the back of the shed, selective or VNA for the fast-moving, high-SKU-count lines near the dock. Splitting the layout by product behaviour, rather than committing the whole floor plate to one system, is usually how a warehouse that grows its SKU count over time avoids re-racking two years in. ## Frequently asked questions ### Can I mix racking types in the same warehouse? Yes, and it is common practice. Bulk, slow-moving stock goes into drive-in bays at the back of the building; fast-moving, high-SKU stock stays in selective or VNA racking near the dock. The split is usually made by pick frequency and SKU count per product line, not by an arbitrary floor-area allocation. ### Does VNA racking need a different building, or can it go into an existing shed? It depends on ceiling height and floor flatness. Turret trucks work best above 30 ft, well beyond most standard-grade sheds, and the taller mast is intolerant of an uneven slab. Retrofitting VNA into a low-ceiling, older-slab building is possible but often costs more in floor remediation than it saves in footprint. ### Is drive-in racking cheaper overall than selective racking? The racking itself is often cheaper per pallet position, and it needs no specialised trucks. But 80-85% effective utilisation is the realistic ceiling once LIFO gaps and SKU diversity are accounted for, so the real saving is smaller than the theoretical density gain suggests, and it disappears entirely if the stock behind it needs FIFO rotation. ### How much more expensive is a VNA turret truck than a reach truck? Sources agree turret trucks cost more to buy and maintain than reach trucks, owing to the rotating head, guidance system and more complex hydraulics, but none quote a reliable AED figure applicable across suppliers and truck classes. Get a landed, maintained cost quote for your specific truck count and lease term before comparing it against the footprint saved. ## The bottom line The layout decision is really a decision about which cost you would rather carry: selective racking carries its cost in unused aisle floor, drive-in carries its cost in retrieval time and LIFO discipline, and VNA carries its cost in equipment and building height. None of the three is wrong in isolation — each is wrong only when it is applied to stock that does not suit it. Before signing a warehouse lease sized around one assumed layout, run the pallet count both ways through the space calculator and price the difference at the rent you have actually been quoted. A layout chosen on paper, before the lease is signed, is far cheaper to change than one discovered to be wrong after the racking is bolted to the floor. *Figures were verified on 6 September 2026 against Hammerhead LLC, NTL Storage, Krost Shelving and Knight Frank's UAE Industrial and Logistics Report H2 2025. Density and cost figures for racking systems vary by supplier, pallet weight and building height, so treat the ranges here as planning inputs rather than a quote.*