Block Stacking: Stack Height, Lane Depth and FIFO Losses

The bottom cartons in lane B-14 start to bow on a humid night. By 02:00 a three-high column has leaned into the aisle, the top pallet is on the floor, and six older pallets behind the front row cannot be reached without moving everything in front of them. The WMS still showed the lane as a tidy 15 positions with room to spare, because nobody had told it how many tiers that carton can actually carry.

Block stacking is the cheapest way to store pallets: no steel, no rack inspection, no beam levels to adjust. The savings disappear when stack height is copied from another product, lanes are dug deeper than the lot size can fill, or receivers top up half-empty lanes with new stock. Three calculations prevent most of that: the load on the bottom case, lane depth from lot size, and the age of whatever sits behind the face of each lane.

• What this guide covers: where floor stacking pays, the crush arithmetic behind a tier limit, ceiling and slab caps, lane-depth and honeycombing math, FIFO losses in single-access lanes, and the WMS rules that keep a floor-stack area honest.
• What it adds: a stack-strength calculation using the McKee box formula, a lane-depth table for one SKU, a carton-upgrade versus floor-rent trade in dollars and rupees, an incident trace and a layout review checklist.

Where block stacking beats racking and where it quietly loses money

Block stacking wins when a site holds few SKUs with many pallets each, loads that can bear weight on top, and stock that turns within weeks. It loses when SKUs are many and lots are small, because every partly empty lane holds floor that no other product may use until the last pallet leaves.

The method itself is simple. Unit loads sit directly on the floor in lanes running back from an aisle, stacked as many tiers as the product allows, and a lift truck works only the front column. Nothing separates the tiers except the pallet and the cartons below it. That means the product's own packaging is the structure, and the lane's only door is at the front.

• Good fit: beverages, bagged or boxed bulk goods and seasonal pre-builds where one SKU fills 40 or more pallet positions and arrives in full lots.
• Poor fit: long-tail SKUs with one or two pallets each, crushable or unstable loads, and products with tight remaining-shelf-life rules at customer receiving.
• Hidden cost: floor held by part-empty lanes, damage from overstacking, and labour spent digging out pallets trapped behind newer stock.

Our guide to pallet racking types compares selective, drive-in, push-back and flow rack by lane access and cost per position. This article stays on the bare floor: how high a stack can go, how deep a lane should be, and what single-door lanes cost in aged stock. Use it when racking has been ruled out, or when deciding whether a block-stack zone should stay at all.

Because the carton is the structure, the limits in a block-stack area are set by several people who rarely meet: the packaging engineer, the structural engineer who designed the slab, the fire protection engineer, and the inventory controller who configures the WMS. The table below names who holds each input and who must sign off before anyone changes it.

Load on the bottom case: the arithmetic behind a safe stack height

The tier limit comes from one number: the weight resting on a single bottom-layer case of the lowest pallet, multiplied by a safety margin and corrected for humidity. Compare that with the carton's compression strength and the answer is either two high, three high, or a board upgrade.

Take an illustrative carton 12 by 10 inches in footprint, 9 inches tall, weighing 14 pounds. Sixteen cartons make a layer on a 48 by 40 inch pallet and five layers make 80 cartons, so the unit load is 1,120 pounds of product plus a 50-pound pallet, 1,170 pounds in all. A bottom-layer case carries its share of the four layers above it in its own pallet: 64 cartons times 14 pounds, divided by 16 columns, or 56 pounds.

• Each added pallet: 1,170 pounds divided by 16 columns, about 73.1 pounds more on every bottom case.
• Two high: 56 plus 73.1, or 129.1 pounds on the bottom case of the lowest pallet.
• Three high: 56 plus 146.2, or 202.3 pounds on that same case.
• Column-stacked assumption: each case sits squarely on the one below. Interlocked patterns, overhang past the pallet edge and broken deck boards concentrate load on fewer corners and raise the real figure.

Stack-strength check for one bottom-layer case (illustrative values)

Load_bottom = (L_own - 1) x C x w_case / n_col + (T - 1) x W_unit / n_col
  L_own  = layers per pallet (5)       C = cases per layer (16)
  w_case = case weight, lb (14)        n_col = columns per layer (16)
  T      = tiers in the stack          W_unit = unit load incl. pallet, lb (1,170)

BCT_required = Load_bottom x SF / R_h
  SF  = safety factor for time under load, misalignment and handling (3.0, set by packaging)
  R_h = strength retained at storage humidity (0.68 at 80% RH, from a 32% loss)

McKee, simplified:  BCT = 5.87 x ECT x sqrt(Z x t)
  ECT = edge crush, lb/in   Z = box perimeter, in (44)   t = board caliper, in

T = 2:  Load 129.1 lb  -> BCT_required 570 lb -> single-wall (t = 0.16) needs ECT 36.6
T = 3:  Load 202.3 lb  -> BCT_required 893 lb -> single-wall needs ECT 57.3
                                               double-wall (t = 0.25) needs ECT 45.8

The McKee relationship and the humidity correction come from a Georgia Tech technical note on box compression, which reports a 32% strength loss at 80% relative humidity in published industry loss factors. The safety factor of 3.0 is illustrative; the packaging engineer should set it from the supplier's own compression tests and the expected days under load, because strength keeps falling while a box carries weight.

Block-stack tier limits for corrugated unit loads come from the load on one bottom-layer case, times a safety factor, divided by strength retained at storage humidity. OSHA 29 CFR 1910.176(b) requires tiered storage limited in height for stability, and a 32% humidity loss at 80% RH can collapse a marginal third tier.

In this example a single-wall board with a tested edge crush of 40 pounds per inch gives a McKee BCT of about 623 pounds: enough for two high in an 80% humidity store, not enough for three. A coastal Indian warehouse in the monsoon and a US Gulf Coast DC in August face the same physics, so the tier field in the item master should be set for the worst month, not the dry season.

Ceiling, slab and sprinkler limits that cap a floor stack

Crush usually stops a block stack long before the roof does, but three other limits still need checking: the clearance below sprinklers, the slab's design load, and the reach of the lift trucks working the lanes. Whichever is lowest sets the tier count in the WMS.

For sprinkler systems installed to meet an OSHA standard, 29 CFR 1910.159(c)(10) sets a minimum 18-inch vertical clearance between sprinklers and the material below. In an illustrative building with deflectors 30 feet above the floor, that leaves 28.5 feet. Three tiers of the example load at 50.5 inches each reach only 12.6 feet, so the building's cube is less than half used.

Sprinkler clearance in a block-stacked warehouse is a separate cap from carton crush strength: OSHA 29 CFR 1910.159(c)(10) requires at least 18 inches between sprinklers and stored material, and the fire design may set a lower storage height. A three-tier stack of 50.5-inch loads reaches only 12.6 feet below a 30-foot deflector line.

That gap tempts planners to stack higher, and the fire design may say no for reasons unrelated to the 18 inches. Storage height, commodity class and the sprinkler type all feed the sprinkler design, and our article on ESFR sprinkler storage height explains why ceiling-only systems cap storage height. Treat the fire engineer's maximum storage height as a separate field from the carton tier limit.

• Slab load: three tiers of 1,170-pound loads put 3,510 pounds on a 48 by 40 inch footprint, about 263 pounds per square foot under the pallet and 239 across the lane including the 4-inch gap. OSHA 29 CFR 1910.22(b) requires each walking-working surface to support its maximum intended load, so the slab figure belongs on the layout drawing.
• Lift height: placing a third tier means lifting a pallet's underside above the top of the second, about 101 inches here plus clearance. Check the mast's lift and the capacity plate at that height.
• Stack stability: a tall column on a narrow base leans further from the same fork bump than a squat one, and block stacks have no beams to catch a leaning load. Stretch-wrap condition matters more on the top tier than the bottom.

Lane depth math: positions, aisle share and honeycombing loss

A deeper lane spends less floor on aisles but leaves more empty positions while a lot is drawn down, because a lane can only take another SKU once it is empty. The right depth balances those two losses for each SKU's lot size and tier count.

Honeycombing is the name for those empty positions. An Auburn University dissertation on block-stacking layouts defines it as unoccupied slots in a partly filled lane that only the SKU already in that lane can use, and notes the core trade: shallow lanes waste less to honeycombing but need more aisles, while deep lanes do the reverse. Our article on space utilization versus cube utilization shows how honeycombing distorts site-level ratios.

• Positions per lane: p = d × h, where d is lane depth in pallets and h is tiers.
• Floor per lane: F = lane width × (d × pallet depth + aisle width / 2).
• Lanes per lot: k = Q / p, rounded up, where Q is pallets per lot.
• Floor per average pallet held: area over the lot's life divided by the average pallets on hand, (Q + 1) / 2.

Those formulas need four inputs: lot size in pallets, tier count, pallet depth into the lane, and aisle width. Each lane is charged its own floor plus half of the aisle in front of it, because lanes on the opposite side share that aisle. If a lot is drawn down steadily and lanes are emptied one at a time, the average number of open lanes over the lot's life is roughly half the lanes it filled, plus half a lane.

Setting the derivative of that area to zero gives a quick estimate of the best depth: d* = √(A × Q / (2 × h × l)), with A the aisle width and l the pallet depth, both in the same unit. Because lanes come in whole pallets and lots split into whole lanes, check the depths either side of d* before painting lines; the rounding often moves the answer by one.

Lane depth for block stacking follows d* = √(A × Q / (2 × h × l)), checked against whole-lane rounding. For a 60-pallet lot stacked three high with 48-inch pallets and a 13-foot aisle, five-deep lanes need about 7.97 square feet per pallet held, against 9.17 for twelve-deep lanes running at 60.5% average occupancy.

Picking a lane depth for each SKU from lot size and stack height

For the example SKU, a lot of 60 pallets stacked three high works best in lanes five deep, at about 8.0 square feet per average pallet held. Lanes of two or twelve deep cost 15% to 20% more floor for the same stock.

The figures below assume a 40-inch pallet face plus a 4-inch gap (a 44-inch lane), 48-inch pallet depth and a 13-foot aisle for counterbalance trucks, all illustrative. With those inputs the estimate gives d* = √(13 × 60 / 24) = 5.7, and the whole-number check confirms five. At two high, the same lot gives d* of 7.0, and six deep turns out marginally better than seven.

Two columns answer different questions. Floor at receipt is what the SKU needs on the day the lot lands, which matters if lanes are dedicated to it. Floor per average pallet held is the right measure when emptied lanes go back into a shared pool. A ten-deep lane looks cheapest at receipt, yet over the lot's life it holds a third of its positions empty.

• Indian metric version: with 1,200 by 1,000 mm pallets in 1.1-metre lanes, a 3.6-metre aisle and lots of 80 pallets, the curve has the same shape.
• Three high in India: six-deep lanes are best at 0.67 square metres per pallet held, matching d* of 6.3.
• Two high in India: eight-deep lanes are best at 0.93 square metres per pallet held, against d* of 7.7.
• Practical answer: most floors carry several SKU families, so set two or three standard lane depths and assign each SKU by lot-size band.

FIFO losses when a single-access lane is topped up with new stock

A block-stack lane is a stack with one door, so within a lane the newest pallet always leaves first. FIFO survives only at lane level, and only if a lane is never topped up with a newer lot before it empties.

The loss mechanism is easy to miss. Suppose lane B-14 holds six pallets of an early-September lot at the back and nine positions free at the front. A receiver putting away a late-September lot sees open positions and fills them. The six old pallets now wait until all nine new ones leave, and if the lane is topped up again before then, the back of the lane never empties at all.

• Waiting time added: pallets placed in front divided by the lane's draw rate. Nine pallets at one pallet a day adds nine days to the age of the six behind them.
• Repeated top-ups: a lane refilled whenever it has room behaves like a stack, and its back column can age without limit.
• Allocation blind spot: the WMS may allocate the oldest lot correctly, but the driver cannot reach it, so the pick is overridden to the front pallet and the record of the override is often lost.

The money is in remaining shelf life. Using illustrative figures, six trapped pallets of 80 cases at a cost of $11.40 a case hold $5,472 of stock. If they age past the customer's acceptance window and go to a liquidator at 30% recovery, the loss is $3,830. The same six pallets in an Indian DC, at 64 cases of ₹420 each, carry ₹161,280 at cost and lose ₹112,896.

Two practical rules keep the loss small. First, size lanes so a lot fills whole lanes and leaves at most one part lane, which the lane-depth table already encourages. Second, record a lane open date in the WMS and flag any lane whose open date is older than the SKU's allowed dwell, so a supervisor clears it deliberately rather than discovering it at a cycle count.

Stronger cartons or more floor: a worked trade in dollars and rupees

A carton upgrade that permits a third tier pays back only when pallets stay long enough for the floor saved to exceed the extra packaging spend. For the example SKU the break-even dwell is about 96 days in the US and 104 days in India.

The comparison uses the best lane depth at each height: two high at six deep against three high at five deep in the US, and two high at eight deep against three high at six deep in India. Floor rent is an illustrative fully loaded occupancy cost; the carton premium is an illustrative quote for moving from single-wall to double-wall board. Replace both with your own figures before deciding.

The answer surprises people who assume taller is always cheaper. For a product that turns in days, floor is cheap relative to packaging bought for every case shipped, so a two-high limit is the right answer. For seasonal stock built months ahead, the stronger board pays, and some sites buy it only for the pre-build run. We found no rule in either market that fixes a numeric floor-stack height, so in India as in the US the choice is commercial once crush, fire and slab limits are met.

A leaning stack on the night shift, traced from putaway to write-off

Most block-stack losses come from two quiet configuration errors meeting on one shift: a tier limit copied from the wrong product and a putaway rule that allows top-ups. The trace below is a composite drawn from common failure patterns, with illustrative document numbers.

1. 21:40, goods receipt GR-4500817: 24 pallets of lot L0920 arrive. The WMS directs nine of them to lane B-14, which still holds six pallets of lot L0902 at the back. The lane's tier field reads 3, copied when the SKU was set up from a heavier-board sister item.
2. 22:05, putaway confirm PA-778201: the driver places the nine pallets three high across the front three columns. The system accepts the move because the lane has nine free positions and no lot-mixing check.
3. 23:30, shift walk: the lead notes corner bulge on the bottom layer of the front-left column. Warehouse humidity reads 82%. No hold is placed because no rule says what bulge means.
4. 02:15, collapse: the front-left column leans about four inches and the top pallet falls into the aisle. Two pallets are booked to damage through adjustment ADJ-DMG-0311. Nobody is hurt; the aisle is closed for 90 minutes.
5. Day 41, customer rejection: the six L0902 pallets, finally reached after the front stock ships, fail the customer's remaining-shelf-life check and go to a liquidator, the $3,830 loss shown earlier.

The controlled version changes two fields. The item master tier limit is set from the crush calculation at 2, and the putaway rule rejects any lane holding a different lot or a receipt date more than two days apart. Lot L0920 then goes to an empty lane, the bulge never starts, and lane B-14 empties on its own schedule.

Lane rules the WMS should enforce and calls the floor lead keeps

Tier counts, lot mixing and lane open dates are data problems that the WMS can enforce without exception. Judging a damaged pallet, a leaning column or a humid week is still a person's job, with software at most helping them notice sooner.

Many WMS products model a block-stack lane as one location with a capacity, which hides the lane's single door. Configure lanes so the system knows depth and tiers separately, so it can tell which pallets are reachable, and keep an open-date attribute per lane. The table splits the work by task.

Keep model outputs advisory. A bulge detector that misses a damaged corner is worse than none if supervisors stop walking the lanes, so the photo rule exists for the person, not the model. The tier field and lot window are different: they should never be overridable from a handheld, because those overrides are exactly how the incident above began.

Lot rotation in block-stacked lanes depends on the WMS putaway rule, not picking discipline: a lane accepting only its original lot until empty keeps first-in, first-out at lane level. Lanes topped up with newer stock act as last-in, first-out; nine pallets placed in front of six add nine days at one pallet daily.

Block-stack layout review before the first lane is painted

Paint is the cheapest part of a block-stack zone and the hardest to change once drivers learn it. Run the checks below with the packaging, structural and fire inputs in hand, and record who approved each one.

Floor zones often sit beside fixed pick faces and reserve racking, and the rules for where each SKU lives interact. Our comparison of fixed versus random storage locations covers how lane assignment rules fit with the rest of the slotting policy. For the wider set of warehouse design and transport topics, see the logistics and supply chain hub.

If you are planning a floor-stack zone, re-laying lanes after a product change, or adding lot and lane rules to an existing WMS, we can review the calculations and configuration with your team. Share your SKU list, lot sizes and building limits through our project request form, and we will come back with a lane plan and the rules needed to hold it.