Batch Size for Machining: Where the Benefit Ends

Why batch size quickly affects money

Batch size in machining changes unit cost almost immediately. The same part made in a batch of 20 and in a batch of 200 will have a different unit price even if material, tooling and program don’t change.

The reason is simple: setup time doesn’t disappear. The workshop spends it once per series and then spreads it over the number of parts. If setup took 2 hours and cutting one part takes 6 minutes, at a batch of 20 each part carries a noticeable share of those 2 hours. At a batch of 200 that share is much smaller.

On paper a large series almost always looks cheaper. But that is true until finished parts start piling up in the warehouse. Money has already gone into material, machining, wages and power, while revenue hasn’t arrived. If the customer takes parts in portions or shifts the schedule, finished goods quickly turn into frozen cash.

There’s another problem. A large batch occupies the machine for a long time. If an urgent order arrives, the plan breaks. The foreman must choose: miss the urgent job or stop the current run. Either way the shop loses money — through penalties, rush work, a new setup or idle time.

Too small a batch also hits profit. When operators switch the machine between parts frequently, spindle hours go into preparation instead of production. Five setups of 40 minutes in a shift are more than 3 hours when the machine isn’t producing finished parts.

Losses usually come from three places: excess setup time, inventory that sits in storage too long, and a disrupted plan when an urgent job gets inserted into a long run.

So increasing batch size without calculation is risky. A low unit price often looks good only in a table. In a real shop you need a wider view: how many hours the machine setup consumes, how many days parts will sit in stock, and how the batch affects CNC load planning. This is where the economics of batch size changes fast.

What changes when the batch grows

When a shop increases batch size, unit cost usually falls at first. Setups occur less often and their time is spread over more parts. But you can’t look only at cutting time. In turning or milling departments a large batch smooths machine loading but makes the cash picture more complex.

The first hidden cost is work in progress. Metal is bought, the machine has done some operations, but money hasn’t returned yet. If washing, inspection, heat treatment or grinding follow machining, inventory between operations grows quickly. On paper you have many parts, but free cash is lower.

Then finished goods inventory grows. That seems safe until demand changes. In reality demand almost always shifts: the client shifts the schedule, an engineer tweaks a dimension, a buyer orders less, or an urgent job gets higher priority. As a result part of the batch sits longer than planned.

A large batch also substitutes real need with setup convenience. The shop produces more because it is easier to run the machine that way, not because the client currently needs the full volume. A month later half the parts may have been made against yesterday’s demand.

This is especially visible on CNC machines in small and medium shops. While one machine runs a long series, small orders queue up. The worst moment is when an urgent job for 10–15 parts arrives and the chuck is already set up for a long run. Stopping the current batch disturbs the rhythm; not stopping delays the urgent order.

Result: a large batch reduces the share of setup in unit price but simultaneously freezes more money in WIP and stock, increases the risk of overproduction and complicates CNC load planning.

A short example. A shop turns bushings in batches of 40 and spends 90 minutes on setup. At 120 pieces the setup cost per part is indeed lower. But if the customer picks up 30 pieces per week, the extra 90 pieces just sit. They occupy space and already contain metal, machine time and operator work. The gain on setup exists, but inventory and loss of flexibility can easily eat it.

A large batch is useful where demand is stable, routing is short and the shop can accept an unscheduled job without trouble. If any of these points don’t hold, increasing the batch quickly stops saving money.

How to calculate a working batch size

You don’t pick a working batch by habit. Calculate it from simple numbers: setup cost, how fast parts leave the shop and how much schedule slack you need to avoid missing urgent jobs.

Too small a batch eats hours in frequent setups. Too large a batch freezes cash in stock and clogs finished goods. You need a size where the machine runs smoothly and parts don’t sit idly.

What data you need

First write down setup time and the machine-hour rate. If a CNC lathe setup takes 2 hours and the machine hour costs 25,000 tenge, one setup costs 50,000 tenge. That is your starting point.

Then look at actual shipments, not the plan. Use weekly or monthly sales for recent periods and remove one-off spikes. If a part normally ships 180 pieces a month but once shipped 500, don’t base the whole calculation on that peak.

The third number is acceptable storage time. For some parts it’s 5–7 days, for others a month. Consider not only shelf space but also risk: do orders change, is rework common, does the customer often request a different version?

Usually five data points suffice: setup time, machine-hour cost, average weekly or monthly shipments, allowable storage time, and schedule reserve for urgent orders.

How to choose a working option

Don’t look for one “perfect” answer. Compare at least three batch sizes — for example, covering 3 days, 7 days and 14 days of demand. For each option calculate how many setups occur per month, how many parts will sit in inventory and how many free hours remain in the schedule.

If a 3-day batch causes too many setups, it’s expensive. If a 14-day batch consumes almost the entire schedule and leaves no window for urgent jobs, it’s inconvenient. The working option is often in the middle.

It’s useful to keep a small window in the plan. Even 10–15% of shift time for urgent items often prevents breakdowns. Otherwise any unscheduled request can ruin the month.

Recalculate batch size when demand, setup time or order mix changes. In shops with short runs and frequent SKU changes, the old calculation becomes obsolete quickly. A good sign is simple: batches get shipped on time, inventory doesn’t bloat and setup crews aren’t running between machines all day.

A simple example for a small shop

A small shop turns a bushing for service orders. Clients usually take 30–40 pieces a week. Demand exists but is uneven: one week 28, another 43, and occasionally an urgent repair appears.

The problem starts with setup. To launch this bushing a master spends about 2 hours: fit fixtures, select tools, make a trial part and check dimensions. If you run batches of 10, those 2 hours fall onto just ten parts. Unit cost quickly rises even though turning takes only minutes.

On paper a large batch looks better. One run of 300 parts noticeably lowers unit cost because setup time is spread across the run. But that solution has its price. If the shop sells on average 35 bushings a week, a stock of 300 sits for almost two months. Money is frozen in finished parts, shelf space is taken, and part of the batch may not move as fast as expected.

There’s a second problem. While the machine runs the large batch, it’s harder to free it for urgent orders. If a more profitable job or an emergency repair arrives, the schedule begins to crack. Either the current batch is stopped or the urgent order is delayed, upsetting the client.

In such cases the working batch often sits above the technical minimum, not where unit price is minimal. For this example a practical range is 80–100 pieces. That volume covers about 2–3 weeks of demand without overfilling the stock.

Why this works in practice: you don’t set up every week, inventory doesn’t sit as dead weight, the schedule keeps a window for urgent jobs, and the master notices demand shifts sooner and avoids unnecessary production.

At 90 pieces the flow is calm: one setup, a few weeks of shipments, then a new run based on remaining stock and order queue. Unit price is much lower than at batches of 10 or 20, but the shop isn’t locked into a long cycle. For small manufacturers this often matters more than squeezing another percent from costs.

When a large batch already hinders

A large series looks attractive if you only consider unit price. But in the shop it often starts to get in the way sooner than calculations show. The machine stays busy, the schedule loses flexibility and any mistake becomes costly at scale.

This is especially visible when sales have promised a fast lead time to a new client. The order exists, material and people are ready, but the required machine is busy with a two-day run. Setup is delayed, the urgent job waits, and the shop loses not hours but the client’s trust.

A too-large batch often collides with changes during production. An engineer corrected the drawing, the customer changed a size, or the technologist found a better operation. If only the first ten pieces were produced losses are tolerable. If two hundred are already made, part of the volume may become unsellable.

Inventory grows quietly but hurts hard. Finished parts sit while shipments lag. Money has gone into material, machine time, tooling and wages without returning. When stock grows faster than shipments, the shop struggles to decide what to run next.

Another issue: small defects hide inside a long series. After replacing an insert dimensions shift, a burr appears, tolerances drift from heating. If inspection is delayed to the end of the run, scrap accumulates unnoticed. Instead of two bad parts you get dozens for rework or write-off.

A large batch usually becomes harmful when several signs align: the machine is occupied longer than the shop can afford, demand for the item is uneven, drawings or specs still change, quality checks are infrequent, and inventory grows faster than shipments.

In practice it’s better to keep time and machine buffers than cling to a long series at any cost. A batch of 40 parts can be more useful than 200: easier to fit into CNC scheduling, simpler to check on first pieces and less risky to stop if the order changes. If a series already prevents taking urgent jobs and inflates stock, the benefit of upsizing is over.

Where mistakes are made most often

Most shops lose money not in cutting but in poor batch calculation. Batch size is often chosen by a single number — the part cost in the machine. That’s not enough. Storage, frozen cash, sorting risk and urgent orders that jump the queue always sit nearby.

The first common mistake is counting only manufacturing cost and ignoring storage. A large batch can reduce setup share per part, but then parts sit for weeks, taking space, requiring tracking, packing and sometimes reinspection. If the part is expensive or made from costly material, the shop’s savings on the machine quickly vanish.

Average demand also deceives. On paper it looks steady: 400 parts a month, so run large batches. In reality demand spikes. One week almost nothing, the next the client needs everything at once. If the machine is tied to a long run, urgent orders queue. Setups drop but delays and stress rise.

Another error is using the same batch size for everything. Convenient in a spreadsheet, harmful on the shop floor. A simple bushing, an expensive housing and a precision shaft require different launch logic. One rule for all almost always causes imbalance.

A fourth mistake appears after process changes. The shop updates tooling, fixtures, machine or routing, but the old batch calculation stays. Yet after such changes you should recalc first. If setup time falls by 20 minutes, the previous large batch may stop being economical. The same happens when a part moves to a different CNC or the operation sequence changes.

A small example. The shop used to set up a lathe in 70 minutes and ran 300 parts. Then a more convenient fixture reduced setup to 35 minutes, but batch size stayed the same. Inventory nearly doubled and urgent jobs on neighboring items began to wait for a free window. Formally nothing was violated, but the calculation was outdated.

Usually the problem shows by several signs: parts made “just in case” pile up, urgent jobs regularly break the shift plan, one batch size is launched across many SKUs, setup norms in calculations are older than actual tooling, and masters often ask to stop a long run unexpectedly.

If your shop changed the machine, tooling or order mix in recent months, recalc batches for the most frequent parts. Even this step often reveals where the shop overpays for inventory and where it creates a queue of urgent work itself.

Quick checks before launching a run

Before starting a large batch take ten minutes and check the decision itself, not the machine. A mistake here costs more than one extra setup. Money goes to stock, the schedule shifts and urgent orders can break the whole plan.

For many shops batch size seems simple: the bigger the run, the cheaper the unit. On paper that’s often true. In practice the saving quickly melts if part of the volume sits in inventory or if an order arrives in three days that can’t be inserted without overtime.

Before a run, ask a few short questions. Will the client buy the whole volume within the next month or later? Will the schedule keep a window for an urgent job without an all-nighter and extra shifts? Is there a risk the drawing, tolerance or material will change? Is there space for this volume and a clear shipment date for each item? And does the calculation produce a noticeable saving, not a few cents per part?

The first question is the dullest but most useful. If demand for the next 3–4 weeks is not confirmed, a large batch turns production into a warehouse. Parts are ready but cash is frozen and shelves fill with items not yet shipped.

The second question is about flexibility. If CNC scheduling is already tight, a large batch can take all breathing room. Then any urgent order can only be inserted at the cost of delaying other items. If you don’t have at least one free window for an unplanned job, reduce the batch.

The third question is often underestimated. In metalworking drawings can be clarified even after approval, especially for a part that goes into a new machine or assembly. If the customer may change material, thread, chamfer or fit, don’t fill the shop with extra volume.

Also look at the numbers. If a larger batch saves only 2–3% due solely to fewer setups, while storage, handling and rework risk eats that difference, upsizing is pointless.

A simple rule: if you’re not confident on at least two questions, reduce the first batch size. For the shop that’s usually calmer and cheaper than later unpacking a pile of finished but unshipped parts.

What to do next

Don’t rewrite the whole production plan at once. Take 2–3 frequently made parts with a track record: setup time, how inventory accumulates and how often urgent jobs interrupt the schedule.

This start gives an honest picture without theory. For one part you’ll see small batches eat the shift in setups. For another, large batches fill shelves and freeze cash. At this stage batch size stops being an opinion and becomes a data decision.

Look at facts from the last month. It helps to make a simple summary per part: minutes spent on setup, how many parts went to stock and how long they stayed, how often urgent jobs cut into the queue, how many hours the machine waited for material, fixture or operator, and how many parts were made outside the normal schedule.

These numbers quickly show where money is lost. Sometimes the issue isn’t batch size but that the shop lives in spikes: large runs one day, then everything stops for two urgent jobs the next.

Then set batch ranges per part group. One uniform rule usually harms. For simple repeat parts ranges can be wider. For parts with frequent changes, manual checks or unstable demand keep batches shorter. If you have many SKUs, start with three groups: stable, seasonal and urgent.

Record limits, not a single number. For example, not always 300 parts, but 180–260 in normal loading and no more than 120 if the week is already full of urgent jobs. That makes it easier for the foreman and planner to decide without long approvals.

If calculations always hit capacity or long setups, the issue isn’t just planning. Then discuss the technological chain: the machine, fixtures, setup time and service. EAST CNC, the official representative of Taizhou Eastern CNC Technology Co., Ltd. in Kazakhstan, works exactly on such tasks: helping pick CNC lathes and machining centers, commissioning and service. Sometimes that’s more useful than endlessly moving batch sizes in a spreadsheet.

A good result looks simple: fewer disruptions to the queue, no bloated inventory, and urgent parts don’t turn a week into a crisis.