How to Test Capacity Before the Next Demand Shift
To test capacity before demand changes, run your plant and each capacity option through several demand scenarios in a simulation: higher volume, a different product mix, a short peak and a ramp-up. Compare every option on the same measures: output, backlog, queue time, overtime hours and where the bottleneck sits. Choose the option that holds up across most scenarios, not the one that looks best on the average week.
Why average-load spreadsheets miss the next demand shift
Most capacity plans start with a sentence like "we can make X units a week." The number comes from a spreadsheet: demand divided by the average cycle time of each station, compared with the available hours. It is a useful first check, and it is also where the surprises come from.
Demand rarely changes only in level. Its shape changes too:
- The mix moves. Volume stays the same, but more orders go to a variant with a longer cycle or more changeovers, and the extra load lands on two or three stations while the rest of the line barely notices.
- Peaks are short and sharp. A few heavy weeks hit the same machines as the base volume, and the backlog takes weeks to clear.
- A ramp outruns people. A new product grows month by month faster than hiring and training.
- Logistics couples everything. More pallets mean more forklift trips, longer queues at the dock and empty buffers in front of busy stations.
In each case the average says the plant has capacity and the floor says it does not. The gap is waiting time, and a spreadsheet has no time axis.
Write the decision as a comparison
Before modeling anything, write the decision in one sentence with real alternatives. For example: "Over the next planning period, do we cover growth with overtime, a third shift on line 2, or a second machine at the paint station?" If there is only one option on the table, there is nothing to test yet.
Typical capacity options worth comparing:
- overtime or weekend shifts
- an extra shift on selected lines
- extra people at one station
- a new machine at the current bottleneck
- a larger buffer or a different batch size
- moving equipment into new or freed floor space
Each option goes into the model as a variant. The demand scenarios stay the same for all of them.
Four demand scenarios to run
The scenarios are stress tests. They show which option breaks first, and where, without anyone having to guess which future will arrive. The numbers below are illustrative.
| Scenario | What changes in the model | What it reveals |
|---|---|---|
| Level shift | Total volume up 15% with the current mix | Whether the planned capacity covers growth at all |
| Mix shift | Same volume, 30% of orders move to a slower variant | Which stations take the extra load and how changeovers grow |
| Peak | Four weeks at 25% above base, then back to normal | How fast the backlog builds and how long it takes to clear |
| Ramp | New product rising month by month while staff are still in training | Whether output keeps pace with orders during the ramp |
Run the current plant first, then each option, under all four scenarios. Use the same shifts, the same nominal cycle times and the same order rules every time. DBR77 Digital Twin runs a deterministic simulation of the week (Averages vs simulation over time), so each scenario gives one repeatable result, and a difference between two options comes from the option, not from noise.
What to measure in every scenario
Keep the list short and identical for all options:
- output in units per shift, and the backlog of orders waiting at the end of each week
- queue time, in minutes, in front of each station and buffer
- which station limits output, and whether that changes between scenarios
- overtime and extra labor hours needed to hold the delivery plan
- forklift and cart utilization, and buffer fill levels
The bottleneck location matters most. If the constraint moves from assembly to painting when the mix shifts, an option that only adds assembly capacity will fail in that scenario. To find that before go-live, see How to Find Production Bottlenecks Before They Happen.
See your capacity options side by side
The demo shows capacity options run through several demand scenarios, with output, queues and the bottleneck position for each. Bring the options you are weighing so we can discuss how to model them.
A real case: adding a hall without stopping production
A car lamp manufacturer in southern Poland, with 270 people on two shifts, needed more capacity. The plant runs injection molding, painting, electroplating, assembly, labeling, quality control and packaging. A new production hall was added, and machines had to be relocated and capacity expanded without interrupting production. The team also had to check intermediate storage capacity and keep intralogistics deliveries running during the moves.
DBR77 built a digital twin that included the new area, with material flows and processing times for each task. Process animations were used to test relocation options and the order of the moves. The results:
- An expansion plan designed to keep output at the target level.
- The main planning challenge was sizing and placing buffer storage between processes.
- Extended transport routes were the main source of production losses.
- The final machine layout differed from the initial assumption.
- The best relocation sequence let some processes run in parallel.
Management reviewed the final design in VR, which helped them choose the sequence of transferred processes faster. New machines and floor space covered only part of the capacity question in this plant. Buffers and transport routes decided whether the added capacity turned into output.
How to read the results and choose
When the runs are done, you usually see one of three patterns:
- One option wins in every scenario. The choice is easy, and you have the evidence to defend it.
- The fastest option is fragile. It wins on the level shift but loses badly when the mix moves or a peak hits. A slower option with smaller losses is often the better buy. We cover this in How to Compare Resilience, Not Just Throughput.
- Several options are close. Then cost, time to implement and how easy it is to undo the change decide. The method in How to Compare CAPEX Options When Every Scenario Looks Plausible helps here.
Say, for illustration, that a second machine at the bottleneck adds the most output in a level shift, while overtime on one shift adds less but costs nothing upfront. Under the mix shift, the bottleneck moves to a station the new machine does not help. Overtime now plus a smaller investment at the second station may then beat the big purchase.
Write down the assumptions that would change the answer, such as a mix share or a ramp speed. When sales updates its forecast, you rerun the same model instead of starting again.
FAQ
How many demand scenarios do we need?
Four is a practical start: level shift, mix shift, peak and ramp. Add more only when a specific risk, such as a large new customer, needs its own test.
Can we use our S&OP forecast as the base?
Yes. Use it as the base scenario and build the others around it. The point is to see how each capacity option behaves when the forecast is wrong in a realistic way.
Do we need live machine data for capacity testing?
No. Routings, nominal cycle times, shift patterns and order history are enough to start. Live data from DBR77 IIoT is a later step, once the model is used regularly.
What if the bottleneck moves between scenarios?
Treat it as a finding. It tells you that a single fix will not cover all demand shapes and that the plan may need two smaller changes.
Conclusion
Start by writing the capacity decision as one sentence with at least two options in it, and ask sales which demand change worries them most. That change becomes your first scenario next to the level shift. If the bottleneck moves between the two runs, look for two smaller fixes before you approve one large purchase.
Planning capacity for next year?
Send us the demand change you expect and the options already on your list. We will say which of them a simulation can tell apart.
Sources
- DBR77, Case study: Plant expansion plan (summary on this site)
- DBR77 Digital Twin, Averages vs simulation over time
Want to test a decision from your plant?
Describe the decision to a DBR77 expert and we will show how DBR77 Digital Twin compares the options on output, waiting time and transport load.
Questions
Have a question about this topic?
Tell us what it is about. We will reply within one business day.
Digital Twin newsletter
Short notes on testing layout and flow decisions on a model, and news from DBR77 Digital Twin. You can unsubscribe at any time.