Transfer Batch vs Process Batch: Cut Lead Time

Transfer batch vs process batch: what is the difference

The process batch (or production lot) is the quantity made at a workstation between two changeovers. The transfer batch is the quantity that moves on to the next operation at one time. The two do not have to be equal, and separating them is one of the cheapest ways to cut lead time.

When a lot of 100 parts leaves machine 1 only after the hundredth part is finished, the first 99 sit waiting and add no value. If the transfer batch is 10 parts, machine 2 starts as soon as the first 10 arrive, and the two operations run at the same time.

Visual map: split the lot to overlap operations

The map shows the difference between the two batches, the formula for total time and the 100-part, three-operation example.

How to calculate the total time of a lot

For N parts, unit times t for each operation and a transfer batch b, with no queues between operations and the setup already done, the time until the last part leaves the last operation is:

T = N × t_bottleneck + b × (Σt − t_bottleneck)

The first term is the time the bottleneck needs to process every part. The second is the time to fill the flow: each non-bottleneck operation is crossed once by one transfer batch. If b = N, the lot moves all together and the formula becomes the serial sum: T = N × Σt.

Example: 100 parts through three operations

Fictional example: a lot of 100 parts goes through cutting (1 min/part), drilling (2 min/part) and deburring (1 min/part). Σt = 4 min and the bottleneck is drilling at 2 min. Setups and queues are left out to isolate the effect of the transfer batch. Handling is assumed at 2 min per move of a batch between two operations.

Going from 100 to 10 parts per transfer drops lead time from 400 to 220 minutes (45%) without changing the machine, the method or the process batch. Going from 10 to 1 gains only 18 minutes and multiplies the moves by ten. The floor on lead time is N × t_bottleneck = 200 min: no transfer batch goes below it.

Changeovers: what stops you from shrinking the process batch

Reducing the transfer batch needs no new changeover, because the machine keeps running the same item. Reducing the process batch does, because every new lot needs a setup. In the example, splitting the lot of 100 into two lots of 50 adds one setup to each of the three operations: at 30 min per setup, that is 3 × 30 = 90 extra minutes of machine downtime.

That is why the usual order of attack is: first shorten the transfer batch, then cut setup time with SMED quick changeover, and only then shrink the process batch. To size the process batch by the balance between setup and inventory, see the economic production quantity (EPQ).

How to choose the transfer batch size

• Natural unit of movement: a bin, a cart, a tray or a pallet layer. A transfer batch that does not fit the container turns into improvisation.
• Move time: compare the time to move with the time to produce the batch. If moving takes almost as long as producing, the batch is too small for that transport method.
• Distance between stations: adjacent stations accept batches of a few parts; stations in different buildings call for larger batches or a supply loop such as the internal milk run.
• Quality risk: parts that move in small batches reveal defects sooner; a fault found at part 10 avoids 90 defective parts.
• Sensitive parts: items that are damaged by handling, such as delicate finishes, may need a larger transfer batch or packaging between operations.

From transfer batch to one-piece flow

Taken to the limit, b = 1 is one-piece flow. It pays off only when stations are close, cycle times are similar and the work between operations has been balanced. Cycle times and customer pace are covered in cycle time, takt time and lead time; the overall effect on delivery time is in lead time.

In practice many plants settle in the middle: they transfer in batches of 5 to 25 parts, according to the container, and keep one-piece flow for cells with fine balancing. To see where material waits today, use value stream mapping (VSM) and measure the queue time between each operation.

Step by step

1. Measure the unit time of each operation and identify the bottleneck (see the theory of constraints).
2. Calculate T for the current lot and for three candidate sizes using the formula above.
3. Define the transfer container and the rule: when the container is full, it moves to the next operation.
4. Set up the work order to release the full process batch but to report and move by transfer batch (see production order).
5. Run for a week, compare measured lead time with the calculated one and adjust the size.

Common mistakes and limitations

• Mixing up the two batches: splitting the order into smaller process batches without cutting setup only increases downtime.
• Forgetting the bottleneck: shrinking the transfer batch in operations before the bottleneck does not raise output; the gain comes from shortening the time to cross the flow.
• Ignoring handling: in the example, b = 1 creates 200 moves. Without a conveyor or adjacent stations, the cost of moving wipes out the gain.
• Batch-type operations: ovens, heat treatment and washers process the whole load at once; there the transfer batch cannot be smaller than the equipment capacity.
• Outdated time data: the formula assumes stable times and no queues. With stoppages and variation, real lead time will be longer than calculated.

Frequently asked questions

What is the difference between a transfer batch and a process batch?

The process batch is the quantity made between two setups; the transfer batch is the quantity that moves to the next operation at one time.

Does reducing the transfer batch need a new setup?

No. The machine keeps the same item; what changes is how often parts are moved.

What is the formula for total time with a transfer batch?

T = N × bottleneck time + b × (sum of times − bottleneck time), with no queues and the setup already done.

In the example, how much does lead time fall with a transfer batch of 10?

It falls from 400 to 220 minutes, a 45% reduction.

When does a one-piece transfer batch not pay off?

When stations are far apart, cycle times differ widely or moving each part costs more than the time saved.

Does a smaller transfer batch increase productivity?

Not directly. It reduces lead time and work in process; capacity is still limited by the bottleneck.

About the author

Vagner Soares

Lean Manufacturing & Behavioral Management Specialist

Over 20 years in the automotive and metalworking industries (GM and Dana), Lean Manufacturing practitioner since 2006. SENAI instructor and mentor in Brazil’s Brasil Mais Produtivo program, delivering consulting, training and audits for 50+ companies, combining quality, productivity and people development.

LinkedInInstagramFull profile →