Central Kitchen Capacity Calculation: How to Size Preparation by SKU, Shift and Peak Demand

Summary

A reliable central kitchen capacity calculation starts with saleable demand by SKU—not with the kilograms-per-hour figure printed on an equipment specification sheet. Planners need to convert outlet orders or customer forecasts into finished kilograms, adjust each SKU for realistic preparation yield, calculate the material load reaching every processing stage, and test whether that load fits inside the actual production window after changeovers, sanitation, breaks, planned maintenance and other scheduled losses.

The same worksheet should translate production volume into labor-hours, equipment runtime, batch counts, work-in-process requirements and peak-demand coverage. This matters because two kitchens producing 3,000 kg per day can require very different configurations: one may run five stable SKUs in long batches, while another handles 40 short runs with different trimming yields, cut specifications and sanitation requirements.

The method in this guide provides a repeatable framework for moving from outlet demand → saleable output → raw input → stage load → shift time → crew and equipment requirement → capacity buffer. It also includes an illustrative three-SKU calculation, reverse checks and sensitivity analysis so planning teams can see where a proposed preparation system fails before equipment or floor space is committed.

Capacity Planning Usually Fails Before Anyone Discusses the Machine

A restaurant group may know that its outlets need three tonnes of prepared vegetables on a busy day. The engineering team may receive quotations showing cutters rated at 1,000 kg/h, washers rated at 1,500 kg/h and dewatering equipment with another stated capacity.

It appears that the calculation should be simple.

Three tonnes divided by several hundred kilograms per hour seems to leave plenty of capacity.

That conclusion can be badly misleading.

The 3,000 kg requirement may refer to finished usable ingredients, while equipment receives raw material before trimming, peeling and inspection losses. Some SKUs may lose 8% during preparation and others more than 20%. A machine may process one product continuously but require blade changes between another three products. Manual trimming may feed the cutter more slowly than its mechanical capacity. A batch dewatering process may become congested even though washing capacity appears adequate.

The production day itself is also shorter than the scheduled shift.

An 8.5-hour shift does not provide 8.5 hours of uninterrupted processing. Start-up checks, breaks, sanitation, product transitions, routine inspections and maintenance consume time. A second shift does not necessarily double output if deep cleaning, cold-room turnover or dispatch schedules occupy the period between shifts.

The right question is therefore not:

“How many kilograms per hour can this machine process?”

It is:

“Can the complete preparation system produce the required quantity of conforming, saleable product within the available time under the expected SKU mix?”

That distinction is the foundation of useful capacity planning.

NIST describes value stream mapping as a way to visualize material and information flows, diagnose constraints and identify opportunities for process improvement. Its process-improvement guidance also treats setup reduction and balanced flow as distinct operating issues rather than assuming that machine speed alone determines output. That logic transfers well to central kitchen capacity modeling: capacity must be calculated through the system, not inferred from one isolated operation.

The Central Kitchen Capacity Calculation Framework

A practical worksheet can be built around seven linked calculations:

1. demand by SKU and demand window;
2. required saleable output;
3. raw-material input after yield;
4. load reaching each processing stage;
5. processing, cleaning and changeover time;
6. labor-hours and equipment requirement;
7. peak-demand and operating buffer.

Each number should be traceable to an assumption or measured plant value.

If a value is unknown, mark it as an assumption and test it through sensitivity analysis rather than disguising uncertainty with false precision.

1. Convert outlet or customer demand into saleable kilograms

Do not begin with raw vegetable purchasing volume.

Begin with what the operation must actually deliver.

For a restaurant network:

Saleable demand = Number of outlets × Required kg per outlet

For portion-based planning:

Saleable demand = Portions required × Ingredient weight per portion

If an ingredient is used in several menu items, aggregate all recipes using the same preparation specification.

A sliced onion SKU and diced onion SKU should normally remain separate because their preparation route, cutting rate, yield and changeover requirement may differ.

The demand window also matters.

A daily average can hide the actual constraint when most production must be completed before a morning dispatch. A facility shipping twice daily may need a six-hour peak production model rather than a 24-hour average.

For planning, build demand at several levels:

• normal operating day;
• peak weekday;
• seasonal or promotional peak;
• expected future volume;
• confirmed new-contract volume.

A peak factor can be calculated from historical orders:

Peak factor = Selected peak demand ÷ Average demand

The appropriate peak definition depends on the business. A company with highly variable institutional orders may choose a different service level from a restaurant chain with predictable replenishment. There is no universal peak factor that every central kitchen should use.

2. Separate firm demand from management reserve

Capacity models often become inflated because several people independently add “safety.”

Sales adds 10%.

Operations adds another 10%.

Engineering adds another 20%.

Purchasing increases raw-material requirements again.

The final project may be sized far above any traceable demand scenario.

A better worksheet keeps the elements visible:

Planned saleable target = Firm demand × (1 + approved service reserve)

The reserve should have an identified reason: order volatility, forecast error, contract service level or another defined business requirement.

Do not hide sanitation loss, machine downtime or poor yield inside the same factor. Those belong elsewhere in the model.

This prevents double counting.

Yield Converts Finished Demand into Raw-Material Requirement

Yield is one of the most commercially important variables in food pre-processing.

For capacity planning:

Yield % = Acceptable saleable output ÷ Raw-material input × 100

Therefore:

Required raw input = Required saleable output ÷ Yield

Suppose a central kitchen must deliver 1,000 kg of trimmed and cut vegetables.

At 90% total yield, it requires approximately 1,111 kg of raw material.

At 75%, it requires approximately 1,333 kg.

The finished requirement has not changed, but receiving, trimming, waste handling, cutting and possibly washing loads have increased substantially.

This is why a single “tonnes per day” figure is insufficient when the SKU mix contains leafy vegetables, root crops, fruits and proteins with different usable yields.

Yield should be recorded by SKU, not by department

A useful SKU sheet separates losses where possible:

• receiving rejection;
• trimming;
• peeling;
• cutting defects;
• damaged product;
• washing loss;
• dewatering or handling loss;
• inspection rejection;
• packing giveaway or reject.

That allows the team to distinguish unavoidable specification loss from process loss.

Over-trimming may improve apparent visual quality while making ingredient cost worse. Aggressive washing may increase damage. Poor cutting consistency can create unusable fragments and may also affect downstream cooking consistency because pieces no longer heat at similar rates.

Yield is therefore both a capacity variable and a quality variable.

Calculate the Load at Every Processing Stage

Raw input is only the beginning.

A cutter does not necessarily receive the full incoming quantity because some mass has already been removed during trimming or peeling. The washer receives material after cutting losses. A dewatering machine may only serve selected leafy SKUs. Packaging receives the final conforming product.

For each stage, calculate the mass that actually reaches it.

One approach is:

Stage load = Raw input × cumulative upstream yield

The same calculation can be performed backward:

Stage load = Required finished output ÷ cumulative downstream yield

The backward method is particularly useful during early project planning because the commercial requirement usually begins with finished output.

Why stage load matters for line balance

Consider two pieces of equipment:

• cutter sustainable processing rate: 700 kg/h;
• washer sustainable processing rate: 900 kg/h.

That does not automatically mean the cutter is the bottleneck.

The cutter may receive 3,000 kg while the washer receives only 2,600 kg after sorting and cutting losses. The cutter may run three product families requiring 40 minutes of total setup, while the washer requires 90 minutes of sanitation transitions. The washer may also need slower operating conditions for delicate products.

Capacity is:

load divided by usable time under the required product conditions.

This is the basis of line balance.

Time Is the Second Denominator

A common worksheet error is to use the full paid shift as machine runtime.

A better model defines a base production window:

Base production window = Shift duration – breaks – start-up – scheduled sanitation – planned maintenance

Product-specific transitions can then be modeled separately:

Stage time required = Processing time + changeover time + other stage-specific planned loss

For continuous equipment:

Processing time = Stage load ÷ sustainable processing rate

For batch equipment:

Batches required = Stage load ÷ effective batch load

Round batch count upward where partial batches cannot be combined.

Then:

Processing time = Batches required × actual batch cycle time

The cycle should include the relevant loading, processing, discharge and reset activities—not only the active machine cycle.

NIST’s setup-reduction guidance defines changeover around the transition between one acceptable production run and the next. That distinction is useful in food processing because changing a blade, recipe, allergen status, hygiene classification or product family can consume real capacity even though no saleable product is being produced.

SKU count can matter more than daily tonnes

Compare two plants producing 4,000 kg per shift.

Plant A

• 4 SKUs;
• approximately 1,000 kg per run;
• limited specification changes.

Plant B

• 32 SKUs;
• average batch approximately 125 kg;
• frequent cutter changes;
• repeated wash-downs;
• multiple packaging formats.

The same daily mass does not represent the same workload.

Plant B may require more operators, staging space and cleaning time despite processing identical total tonnage.

A useful KPI is:

Changeover burden = Total changeover minutes ÷ Base production window

Another is:

Average campaign size = Total SKU output ÷ Number of production campaigns

These make product complexity visible.

Build Labor Capacity from Labor-Hours, Not Headcount Alone

Central kitchen labor is often discussed as “20 people today versus 10 people after automation.”

That is too crude for planning.

Different preparation tasks consume different labor-hours.

For each SKU and activity:

Labor-hours = Stage load × Standard labor minutes per kg ÷ 60

Add fixed tasks such as:

• line setup;
• sanitation;
• QA checks;
• internal transfer;
• crate handling;
• documentation;
• waste removal.

Then:

Base crew requirement = Total required labor-hours ÷ Effective hours available per worker

If historical attendance, training, leave and absence data show that the nominal roster is not consistently available, a site-specific relief factor can be applied.

Do not borrow a generic absence percentage from another country or factory.

Use the operation’s own records.

Labor planning matters even when equipment is installed. Automated cutters still need feeding, inspection and cleaning. Continuous washing systems require supervision and water-control tasks. Packaging may require more labor when upstream equipment raises throughput faster than the downstream area can absorb it.

U.S. Bureau of Labor Statistics data illustrates why workforce assumptions deserve explicit treatment rather than being treated as unlimited supply. Its Food Preparation Workers profile reported about 902,700 U.S. jobs in 2024, a May 2024 median hourly wage of $16.45 and roughly 148,000 projected openings per year during 2024–2034, primarily from replacement needs. BLS also notes that food-preparation employees commonly learn through several weeks of on-the-job training. These figures describe the U.S. occupation overall; they are not a staffing or wage forecast for any individual central kitchen or another country.

Central Kitchen Capacity Worksheet: Input and Assumption Table

The following structure can be recreated in a spreadsheet or planning system.

Do not force every field into a single percentage called “efficiency.” Keeping the assumptions separate makes the model explainable and easier to audit.

Worked Hypothetical Example: 24 Outlets and Three RTC Vegetable SKUs

The following is an illustrative calculation only. It is not a TEEMYEAH customer result, equipment commitment or performance claim.

Assume a centralized preparation facility supplies 24 outlets with three chilled ready-to-cook vegetable SKUs during a peak day.

Hypothetical demand and yield assumptions

Total firm demand is:

1,320 + 840 + 672 = 2,832 kg

After the illustrative 3% dispatch reserve:

Planned saleable target = 2,916.96 kg

Applying the three SKU-specific cumulative yields gives approximately:

Required raw input = 3,583 kg

The weighted mass difference between raw input and planned finished output is therefore approximately 666 kg in this hypothetical model.

That difference should not automatically be classified as “waste.” It can include peel, roots, damaged leaves, trimming required by specification, cutting fragments and other losses. A real factory should measure these categories separately.

Convert the model into stage loads

Using the illustrative stage yields:

• total feed reaching the cutting stage ≈ 3,125 kg;
• total feed reaching washing ≈ 3,062 kg;
• leafy material requiring dewatering ≈ 2,337 kg;
• packed saleable target ≈ 2,917 kg.

Now the equipment requirement can be checked against these loads instead of the original 2,832 kg order figure.

Hypothetical stage-load capacity decision table

Assume an 8.5-hour shift with scheduled breaks, start-up, end cleaning and planned inspection/maintenance leaving 6.5 hours of base equipment production window.

The rates below are hypothetical planning values and do not represent TEEMYEAH machine specifications.

The washer is not the bottleneck even though it handles substantial mass.

Dewatering has the lowest spare time.

That becomes important when demand increases.

Calculate the illustrative trimming crew

Using the labor assumptions above:

• cabbage trimming: approximately 15.2 labor-hours;
• lettuce trimming: approximately 15.2 labor-hours;
• carrot preparation: approximately 4.6 labor-hours.

Total:

≈ 35 labor-hours

If a trimming employee provides 7.25 effective direct-production hours during the shift:

35 ÷ 7.25 = 4.83 workers

Five workers are theoretically enough under these assumptions, but the result contains almost no protection against absence, raw-material difficulty, rework or slow batches.

A six-person planned crew creates more operating resilience.

If one of those six workers is absent, five workers provide:

5 × 7.25 = 36.25 labor-hours

The process remains mathematically feasible but becomes tight.

This is a much more useful staffing conversation than saying that the preparation department “normally has six people.”

Reverse Checks: Make the Worksheet Prove Itself

A capacity model should be checked backward before it is used to specify equipment.

Reverse check 1: finished orders

Sum the planned packed quantity by outlet, SKU and delivery wave.

Does it reconcile with the commercial order file?

If not, the model is solving the wrong demand requirement.

Reverse check 2: raw purchasing

Convert calculated raw kilograms into the purchasing unit actually used—crates, bags, cases or pallets.

Can receiving physically supply this quantity within the required time?

If the worksheet requires 3.6 tonnes but purchasing schedules only 3.2 tonnes, a faster production line cannot close the gap.

Reverse check 3: mass balance

For each SKU:

Raw input ≈ saleable output + measured losses + rejects + recoverable rework

Repeated unexplained gaps indicate a measurement problem.

Reverse check 4: time balance

For every stage:

Processing + changeover + planned stage loss ≤ available window

One failing stage is enough to invalidate the line output.

Reverse check 5: labor balance

Compare required labor-hours with the effective hours of employees who are actually trained and assigned to that stage.

Headcount elsewhere in the building does not create preparation capacity.

Reverse check 6: WIP and floor space

If cutting can temporarily produce faster than washing, where does the intermediate product wait?

A useful measure is:

WIP coverage time = Buffer quantity ÷ downstream consumption rate

The result should be evaluated against temperature control, food-safety requirements, product damage risk and available floor space—not simply against how many plastic bins fit beside the machine.

Reverse check 7: dispatch timing

A production system may have sufficient daily capacity and still fail service if the final 1,500 kg is completed after the delivery vehicle must leave.

Capacity must therefore be checked against completion deadlines, not only total shift output.

Sensitivity Analysis: Find the Assumption That Breaks the Plan

A good central kitchen capacity calculation should deliberately stress the model.

The objective is not to predict one perfect number. It is to understand which assumptions matter enough to change the equipment or staffing decision.

Using the hypothetical model above:

This reveals something that a simple equipment list would miss.

The base line is balanced reasonably well. A 15% demand increase is still workable, but dewatering has little recovery margin. At 25% additional demand, adding more cutting capacity would not solve the problem. The constraint has moved downstream.

That is the purpose of sensitivity analysis.

Variables worth stress-testing

Run the model against:

• peak demand;
• lower-than-normal raw-material yield;
• seasonal changes in vegetable size or maturity;
• one employee absent from a labor-intensive stage;
• additional SKU changeovers;
• slower delicate-product settings;
• unplanned short stops;
• longer sanitation;
• delayed raw-material release;
• packing slowdown;
• new customer specifications.

Seasonality deserves particular attention. The same cutter or peeler can experience different feeding and yield conditions when agricultural raw materials vary in size, shape, maturity, soil load or physical strength.

Do not represent that uncertainty with a single unrealistic decimal point.

Use scenarios.

RTE Capacity Cannot Be Calculated as RTC Capacity Plus More Washing

Ready-to-cook and ready-to-eat products may use similar equipment categories, but their capacity models should not automatically be combined.

RTE fresh-cut products can be consumed without a subsequent cooking step. FDA’s fresh-cut produce guidance highlights the absence of a lethal process, the possibility of contamination spreading through handling and product mixing, and the need to manage processing water, sanitation, equipment cleanability and separation between raw and finished product areas. FDA also notes that processing water can spread contaminants if it is not adequately controlled.

That has a direct capacity implication.

Suppose a shared preparation area can mechanically switch between RTC cabbage and RTE salad in 15 minutes.

The mechanical changeover time is not necessarily the hygiene changeover time.

The valid production schedule may need to include:

• product evacuation;
• removal of residual plant material;
• cleaning;
• sanitizing;
• inspection or release;
• tool and container replacement;
• water-system reset;
• concentration checks where antimicrobial wash systems are used;
• personnel or zone-control requirements.

FDA recommends equipment for fresh-cut operations be designed for cleaning and maintenance and recommends raw and finished product separation to reduce cross-contamination potential.

These controls are not “lost capacity” to be engineered away. They are process requirements.

A planner should therefore create separate capacity assumptions for RTE and RTC product families unless the facility’s hazard analysis, sanitation procedures, layout and validated operating method support another approach.

FDA’s 2022 Food Code is a model code for retail and foodservice operations rather than a universal manufacturing regulation, so local jurisdictional requirements still need to be checked.

For U.S. meat and poultry establishments under FSIS jurisdiction, HACCP validation has a specific regulatory context. The broader planning lesson is still useful: critical process assumptions should be supported and checked with actual operating data rather than assumed from design intent alone.

Capacity Buffers Should Protect the Constraint, Not Inflate Every Machine

It is tempting to buy every stage substantially larger than current demand.

That may create unnecessary capital expenditure without improving system performance.

A better capacity discussion asks:

• Which stage is difficult to expand later?
• Which stage experiences the highest product variability?
• Which equipment can process several SKUs?
• Where can production be extended into another shift?
• Which stage has long sanitation or setup times?
• Which stage failure would stop the complete process?
• Can temporary labor cover peaks more economically than permanent equipment?
• Is additional WIP acceptable?
• Is a parallel machine more resilient than one oversized machine?

A central kitchen may intentionally install excess cutting capacity because interchangeable cutters provide flexibility and redundancy while using a single appropriately sized continuous washer.

Another facility may need parallel dewatering equipment because a batch operation is the real bottleneck.

The correct buffer is constraint-specific.

Labor, Automation and Equipment Should Be Compared as Capacity Configurations

Automation should not be treated as the default answer to every labor problem.

The worksheet can compare three operating models.

When manual preparation is still rational

Manual processing can remain commercially sensible when:

• daily volume is low;
• products change constantly;
• individual batches are very small;
• raw material is highly irregular;
• skilled judgment is required;
• equipment cleaning would take longer than the production campaign;
• capital is better used elsewhere.

Employees can respond to variability that fixed equipment may handle poorly.

When semi-automation often provides a better answer

Semi-automatic systems are useful when one or two processes dominate labor or constrain throughput.

Examples include:

• automated cutting after manual trimming;
• mechanical peeling with manual inspection;
• continuous washing fed by manual sorting;
• centrifugal dewatering after manual washing;
• conveyors linking selected high-volume stages.

This approach can preserve flexibility without forcing dozens of low-volume SKUs through one continuous line.

When an integrated line becomes more logical

Greater integration becomes easier to justify when:

• production volume is consistently high;
• a limited number of product families represent most throughput;
• specifications are stable;
• dispatch deadlines create repeatable peak loads;
• manual transfer causes WIP accumulation;
• labor availability constrains expansion;
• multiple machines can be balanced around the same demand model;
• hygiene zoning and sanitation can support the process.

Even then, automation creates new dependencies.

A failed conveyor can stop several machines. Incorrect sequencing may create upstream accumulation. Maintenance capability and spare-parts planning become more important. Operators require technical training as well as food-safety training.

Equipment cannot replace process design, preventive maintenance, sanitation management or competent supervision.

Do Not Ignore Ergonomics When Calculating Manual Capacity

A time study can show that a worker trims a particular vegetable quickly.

That does not mean eight uninterrupted hours at the same pace should be entered into the worksheet.

Food-preparation work can involve prolonged standing, repetitive cutting, wet floors, lifting crates and fast production conditions. BLS identifies standing, lifting, cuts, slips and related hazards within the work environment of food-preparation workers.

Real labor standards should therefore come from representative observation under a sustainable work method.

Potential improvements may include:

• appropriate work-table height;
• shorter reach distances;
• mechanical lifting or tipping;
• conveyors instead of repeated crate carrying;
• job rotation;
• better knife and tool management;
• slip-resistant floor management;
• planned breaks.

A process that only meets output by requiring employees to operate at an unsustainable pace is not genuinely capacity-feasible.

Cleaning Time Is Part of Capacity

Sanitary design affects more than food safety.

It affects how much of the shift remains available for production.

Equipment with difficult-to-access belts, tanks, guards or product-contact areas may require longer cleaning and inspection. Residue-retaining areas can also make sanitation verification harder.

For a mixed-SKU facility, record:

Cleaning minutes per production day

Changeover cleaning minutes per transition

Full sanitation minutes per shift

Cleaning labor-hours per shift

FDA’s fresh-cut guidance recommends easy-to-clean equipment and emphasizes sanitation, water management, records and prevention of cross-contamination. It also recommends records covering areas such as sanitation, equipment monitoring, water treatment, temperature control and processing batches.

These requirements reinforce an important planning principle:

Cleaning cannot be removed from the capacity equation merely because it produces no kilograms.

Reliability Should Be Modeled Once, Not Hidden Everywhere

Another common error is repeated derating.

For example:

• supplier capacity × 80%;
• then shift efficiency × 85%;
• then another 20% safety reduction;
• followed by 20% spare equipment capacity.

The model becomes impossible to interpret.

Where representative production data exist, use an observed or demonstrated sustainable rate that already reflects the agreed operating condition.

If only provisional equipment data are available, state that limitation.

Create scenarios such as:

• expected operating rate;
• conservative operating rate;
• peak-case rate.

For reliability:

Availability = Scheduled available time – unplanned downtime ÷ Scheduled available time

Use actual maintenance records when an existing line provides them.

For a new facility, keep the assumed availability visible and test different values rather than embedding the assumption invisibly inside every calculation.

This article is about sizing the production system. Detailed supplier capacity testing and FAT acceptance should remain a separate verification exercise.

Measure Preparation Cost per Saleable Kilogram

Capacity alone does not decide whether automation is commercially justified.

A useful comparison metric is:

Preparation cost/kg = Total relevant preparation cost ÷ Saleable prepared kg

Relevant cost categories can include:

• direct labor;
• supervision;
• overtime;
• utilities;
• wash-water treatment;
• cleaning chemicals;
• consumables;
• blades;
• preventive maintenance;
• spare parts;
• waste handling;
• depreciation or financing where appropriate.

Yield should be incorporated carefully.

If a new process reduces direct labor but removes more edible product, the apparent labor saving may be offset by higher raw-material cost.

For investment comparison:

Annual net operating benefit = Avoided current cost + incremental contribution from usable additional capacity – additional operating cost

A simple payback indicator is:

Payback period = Project investment ÷ Annual net operating benefit

Payback should be calculated using the company’s actual labor rates, raw-material values, maintenance cost, energy prices, demand forecast and financing assumptions.

There is no defensible universal payback period for a central kitchen processing line.

Turn the Worksheet into an Operating Planning Tool

The model becomes more valuable after commissioning.

Compare planned assumptions with actual performance every month or production cycle.

Track at least:

• kilograms ordered by SKU;
• raw kilograms issued;
• accepted finished kilograms;
• yield by SKU;
• labor-hours;
• equipment runtime;
• changeover time;
• sanitation time;
• downtime;
• rework or rejection;
• dispatch completion;
• WIP accumulation.

That creates a feedback loop.

Perhaps the original model assumed 80% lettuce yield but the actual operation repeatedly achieves 74%. The issue may be raw-material quality, trimming specification, worker practice or damage.

Perhaps the cutter has ample capacity but spends 90 minutes waiting for hand trimming.

Perhaps one outlet’s promotion has shifted demand toward a SKU that consumes twice as much dewatering time as the original mix.

Capacity planning is therefore not a one-time engineering spreadsheet.

It becomes a way to connect sales forecasts, purchasing, production scheduling, labor planning and future equipment selection.

What a Planning Team Should Know Before Requesting a Line Configuration

By the time a project reaches preliminary equipment selection, the capacity worksheet should be able to answer:

• What are the main SKUs?
• What finished quantity is required for each SKU?
• What is normal versus peak demand?
• How much raw material must enter?
• What yield has been measured or assumed?
• Which processes does each SKU require?
• Which stages are shared?
• Which products are RTC and which are RTE?
• What cut specifications are required?
• How many campaigns run per shift?
• How long do changeovers take?
• What sanitation transitions are required?
• How many productive hours are actually available?
• What labor-hours are required?
• Which stage becomes the bottleneck at peak demand?
• How much future expansion is already commercially justified?

Once those questions are answered, equipment selection becomes much more precise.

The project team can compare one larger machine, parallel machines, semi-automation, phased expansion or a complete line against the same demand model.

That is more useful than asking several suppliers for “a 3-ton-per-day vegetable line” and attempting to compare quotations afterward.

Capacity Planning Is a Process-Design Exercise

A central kitchen with sufficient nominal machine capacity can still miss orders because its yield assumptions are wrong, too many SKUs compete for the same equipment, manual preparation cannot feed the line, sanitation occupies more time than expected or one downstream batch process becomes saturated during peak production.

A robust calculation works backward from demand.

Demand defines saleable output.

Saleable output and yield define raw input.

Raw input and route yields define stage loads.

Stage loads, rates and changeovers define required time.

Required time and labor standards define crew and equipment needs.

Sensitivity analysis reveals where the configuration stops working.

Only after that chain is visible should the project move into final equipment matching and layout design.

Automation can improve a well-designed process. It cannot compensate for a capacity model that never represented the real production requirement.

Use the Central Kitchen Capacity Worksheet

Before comparing equipment, prepare one worksheet covering your main SKUs, outlet or customer demand, finished specifications, measured or expected yield, peak production requirement, shift schedule, current staffing and available production hours.

The result can be used to identify which stage needs additional capacity, which stages can remain manual, where semi-automation is sufficient and whether an integrated line is commercially justified.

For projects moving from calculation into configuration, TEEMYEAH supports central kitchens, food manufacturers, fresh-cut processors and RTC/RTE facilities as a food pre-processing equipment supplier, customized food preparation solution provider and vegetable processing equipment manufacturer. The next discussion can focus on preliminary process recommendations, capacity matching, equipment configuration, raw-material testing and layout rather than starting from an undefined tonnes-per-hour request.

Project teams can continue with the central kitchen food processing solution framework or evaluate a custom food pre-processing line once SKU and capacity inputs are sufficiently defined.

FAQ

1. What should a central kitchen capacity calculation include?

A central kitchen capacity calculation should begin with saleable demand by SKU and convert that demand into raw-material input using measured or realistic yield. It should then calculate the load entering each preparation stage, sustainable processing time, batch requirements, SKU changeovers, sanitation time, labor-hours, equipment requirement and peak-demand coverage. Average daily kilograms alone are not sufficient. The model should also show which assumptions are measured and which are provisional so planners can test lower yield, higher demand, worker absence, slower production or additional changeovers before committing to equipment.

2. How do you calculate raw-material requirements from finished output?

Divide required saleable output by cumulative preparation yield. If 1,000 kg of finished product is required and cumulative yield is 80%, the calculated raw requirement is 1,250 kg. Yield should normally be measured by SKU because trimming, peeling, cutting and inspection losses vary by ingredient and specification. When historical data are unavailable, use a clearly marked provisional assumption and run sensitivity scenarios. Do not describe an assumed yield as a guaranteed operating result, especially when agricultural raw materials vary seasonally.

3. Should central kitchen capacity be based on average demand or peak demand?

Both should be modeled, but equipment and staffing decisions need to demonstrate how the operation will handle the demand window that matters commercially. Average volume is useful for utilization and cost analysis, while peak-day or peak-dispatch demand tests service capability. Historical order data can be used to calculate a peak factor, but the chosen service level depends on the business. Avoid applying an arbitrary universal safety percentage. Keep the forecast, peak factor and management reserve as separate assumptions so capacity is not accidentally buffered several times.

4. How does SKU count affect food preparation capacity?

SKU count affects capacity through campaign size, setup, blade changes, sanitation, packaging changes, staging and scheduling complexity. A line producing 4,000 kg across four long campaigns can have substantially more production time available than a line producing the same mass across 30 short campaigns. Record the number of product transitions and the time from the last acceptable output of one run to the first acceptable output of the next. Small batches can make changeover performance more important than maximum mechanical throughput.

5. How should labor requirements be calculated for a central kitchen?

Convert each manual activity into labor-hours rather than estimating staffing from headcount alone. Multiply the applicable stage load by observed standard labor minutes per kilogram, add fixed duties such as setup, sanitation, QA and material handling, and divide the total by effective working hours available per trained employee. Historical attendance and absence data can be used to calculate a site-specific relief requirement. The time standard should represent safe, sustainable work rather than the fastest short demonstration observed during a time study.

6. What is the difference between machine capacity and central kitchen capacity?

Machine capacity describes one piece of equipment under stated or assumed conditions. Central kitchen capacity is the amount of conforming finished product the complete preparation system can deliver within the required production and dispatch window. It depends on raw-material yield, labor feeding, upstream and downstream stages, product changeovers, sanitation, batch cycles, equipment availability, packing and other process constraints. A high-capacity cutter therefore does not establish line output if washing, inspection, dewatering or packing cannot absorb the material it produces.

7. How much spare capacity should a central kitchen have?

There is no universal spare-capacity percentage suitable for every project. The requirement depends on demand volatility, equipment redundancy, production shifts, maintenance strategy, expansion plans and the commercial consequence of missing orders. Instead of increasing every machine by the same factor, identify the stage most vulnerable to peaks and test the complete model at higher demand, lower yield and reduced availability. Some plants benefit from parallel equipment at a critical batch stage, while others can cover peaks through longer campaigns or an additional shift.

8. Can RTE and RTC products share the same capacity model?

They can appear in the same planning workbook, but their production routes, hygiene assumptions and available operating time should not automatically be identical. Ready-to-eat products may reach consumers without a subsequent cooking step, so sanitation transitions, processing-water control, equipment cleanability, protected post-wash handling and hygienic zoning can impose additional time and operating constraints. A mechanically possible product change does not prove that the hygiene changeover is acceptable. The production schedule should reflect the facility’s hazard analysis and applicable regulatory requirements.

9. When is semi-automation more appropriate than a complete processing line?

Semi-automation is often appropriate when a central kitchen has moderate volume, a broad SKU mix or one clearly identifiable bottleneck. A processor might mechanize cutting, peeling, washing or dewatering while retaining manual trimming and inspection for variable products. This can improve the constrained stage without forcing low-volume SKUs through an inflexible continuous line. Full integration becomes more attractive when demand, specifications, product families, sanitation methods and production schedules are sufficiently stable to keep connected equipment productively utilized.

10. What information should be sent to an equipment supplier for capacity matching?

Provide the SKU list, raw-material form, required finished specification, normal and peak output by SKU, measured or assumed yield, RTE or RTC classification, operating shifts, production window, current labor arrangement, changeover requirements and available floor space. Product photos, process videos and representative raw-material samples can further improve equipment matching. For a new project, clearly mark forecast values and assumptions. A useful supplier discussion should convert these inputs into stage loads and operating requirements rather than simply selecting machines whose catalog capacities exceed total daily tonnage.

Data and Source Note

The calculation figures and the 24-outlet worked example in this article are illustrative/hypothetical and are not TEEMYEAH customer results. Rates, yields, labor standards, reserve percentages and shift assumptions must be replaced with project-specific data before engineering or investment decisions are made.

• National Institute of Standards and Technology (NIST), “Value Stream Mapping,” current online manufacturing guidance, accessed 2026. Applicable to manufacturing process and information-flow mapping, waste identification and improvement planning. Used here as methodological support for mapping capacity across connected stages rather than treating individual equipment in isolation.
• NIST Manufacturing Extension Partnership, “Lean and Process Improvement,” 2020. Applicable to manufacturing improvement methods including value stream mapping, setup reduction and flow manufacturing. Used to support the treatment of changeover and balanced flow as independent capacity variables.
• U.S. Food and Drug Administration, “Food Code 2022,” 2022, current 2022 edition version posted January 2023 with later supplement corrections. A model code for retail and foodservice safety in the United States rather than a universal manufacturing regulation. Local adoption and jurisdiction-specific requirements must be verified.
• U.S. Food and Drug Administration, “Guidance for Industry: Guide to Minimize Microbial Food Safety Hazards of Fresh-cut Fruits and Vegetables,” 2008. Applicable to fresh-cut produce processors, including guidance on sanitation, equipment, processing water, product flow, temperature management and records. Particularly relevant when modeling RTE or fresh-cut capacity where sanitation and post-cut contamination controls affect usable production time.
• U.S. Bureau of Labor Statistics, “Food Preparation Workers,” Occupational Outlook Handbook, updated 2025 using 2024 employment and wage data and 2024–2034 projections. Applicable to the U.S. food-preparation occupation overall. It provides labor-market, training, work-environment and wage context but should not be applied directly to staffing costs or labor availability in an individual factory or another country.
• USDA Food Safety and Inspection Service, “FSIS Compliance Guideline: HACCP Systems Validation,” April 2015. Applicable to HACCP validation requirements for establishments under relevant FSIS jurisdiction, particularly meat and poultry operations. It is referenced here for the principle that safety-critical process assumptions need documented support and in-plant verification; it should not be interpreted as a universal regulatory requirement for every central kitchen.