Hierarchical Capacity Planning Methods

Introduction

Capacity planning is the process of determining whether an organisation has sufficient resources to produce the required output at the required time. In production systems, capacity is not a single number. It is distributed across machines, labour groups, tooling, suppliers, inspection facilities, material-handling systems and supporting functions. A production plan may therefore appear feasible in aggregate while being impossible at a particular work centre.

Hierarchical capacity planning addresses this problem by matching the precision of analysis to the planning horizon and the quality of available data. Long-range planning uses broad resource categories and product-family information. Medium-range planning tests key production resources against the master production schedule. Short-range planning uses detailed planned orders, routings, operation times and work-centre calendars.

The central managerial challenge is to obtain useful warning of capacity problems without spending the time and computational effort required for detailed analysis too early. This chapter compares three principal methods: Resource Requirements Planning (RRP), Rough-Cut Capacity Planning (RCCP), and Capacity Requirements Planning (CRP). It also examines the overall factors, capacity bills and resource-profile techniques used in RCCP.

Learning objectives

After studying this chapter, the reader should be able to:

1. Explain the logic and purpose of hierarchical capacity planning.
2. Distinguish RRP, RCCP and detailed CRP by horizon, data, precision and decision use.
3. Calculate capacity requirements, available capacity, utilisation and load profiles.
4. Apply overall factors, capacity bills and resource profiles.
5. Diagnose overloads and evaluate corrective actions.
6. Select an appropriate planning method when data, time and uncertainty vary.
7. Critically assess the assumptions and limitations of capacity-planning models.

1. The hierarchy of capacity decisions

A production planning system normally operates through linked levels:

The hierarchy is not merely a sequence of reports. It is a control structure. Decisions made at an aggregate level constrain decisions at lower levels, while feedback from lower levels may require revisions to the master schedule or aggregate plan.

A useful principle is: do not use more detail than the decision requires, but do not use less detail than the decision can tolerate. RRP is efficient for strategic alternatives but may conceal bottlenecks. Detailed CRP can identify operation-level conflicts but is expensive and highly dependent on accurate data.

2. Fundamental capacity concepts

2.1 Capacity, load and available time

Capacity is the maximum feasible output or processing time of a resource under specified conditions. It may be expressed in units, labour-hours, machine-hours, tonnes, batches or another appropriate measure. The choice of unit must reflect the resource constraint.

For a work centre, nominal available time over a planning period is often estimated as:

where:

• A_n = nominal available time, usually machine-hours or labour-hours;
• N = number of parallel resources;
• S = shifts per day;
• H = scheduled hours per shift;
• D = working days in the period.

This is a gross figure. If planned downtime, breaks, maintenance or absenteeism are excluded, effective capacity can be represented as:

where A_e is effective available capacity and α is the combined availability factor, between 0 and 1. If performance and quality losses are relevant, an overall equipment effectiveness approach may be used:

Here, the three factors are proportions rather than percentages. This expression is useful for operational diagnosis, but it should not be confused with a guaranteed output forecast: the factors may vary by product, shift and operating condition.

Load is the amount of capacity demanded by planned work. A simple utilisation measure is:

where L is required load and A_e is effective available capacity. A utilisation above 100% indicates an overload under the stated assumptions. Utilisation below 100% does not automatically imply that delivery dates are safe, because queueing, due-date concentration, setup dependencies and material shortages may still cause delay.

2.2 Capacity requirements

For an operation, the required time should normally include both setup and run time:

where:

• T_r = total required time;
• Q = quantity processed;
• t_r = run time per unit;
• n_s = number of setups;
• t_s = setup time per setup.

All time variables must use consistent units. If a routing states hours per 100 units, it must be converted before multiplication by individual units. Scrap, yield and rework may require an adjustment to the quantity launched rather than the quantity demanded.

3. Resource Requirements Planning

RRP is the broadest capacity-planning method. It converts an aggregate production plan into requirements for major resource categories such as direct labour groups, machine families, floor space, warehouse capacity or critical suppliers. It is particularly appropriate when products are still represented as families and detailed bills of material and routings are unavailable or unstable.

3.1 Logic of RRP

The aggregate plan specifies production volumes by period. These volumes are translated into resource requirements using historical standards, resource consumption rates or management estimates. The resulting requirements are compared with broad capacity availability.

For a resource category, the basic relationship is:

where:

• R_r,t = requirement for resource r in period t;
• P_t = aggregate production quantity in period t;
• c_r = average resource consumption per unit of aggregate output.

If multiple product families are included:

where P_i,t is the planned quantity of family i and c_i,r is its average consumption of resource r.

RRP is valuable for decisions such as adding a shift, outsourcing, recruiting, acquiring major equipment or revising the aggregate production plan. Its weakness is aggregation. An average labour-hour rate may conceal a specialised machine that is required by only one family.

3.2 Example: RRP

An aggregate plan requires 4,000 units in April. Historical data indicate an average requirement of 0.18 labour-hours and 0.06 machine-hours per unit. The aggregate requirements are therefore 720 labour-hours and 240 machine-hours. If the available values are 800 and 220 hours respectively, labour has 90% utilisation, while the machine group has 109.1% utilisation. The plan is therefore not feasible without overtime, subcontracting, rescheduling or an increase in machine capacity.

The example illustrates both the usefulness and limitation of RRP: it detects a broad shortage, but it does not identify which products, operations or dates create the shortage.

4. Rough-Cut Capacity Planning

RCCP tests whether the master production schedule is feasible with respect to selected critical resources. The MPS is more specific than the aggregate plan because it identifies end items and their timing. RCCP remains deliberately simplified: it usually examines key work centres, labour groups, suppliers or bottleneck resources rather than every operation.

RCCP supports the principle that the MPS should not be accepted as realistic until its major capacity implications have been tested. If a critical resource is overloaded, the planner can alter quantities, timing, product mix, lot sizes, overtime or sourcing before detailed MRP and CRP generate a large number of infeasible orders.

Three classical RCCP approaches are the overall factors method, the capacity bills method and the resource-profile method.

5. Overall factors method

The overall factors method estimates capacity requirements by applying an average capacity factor to the MPS quantity. It is the simplest RCCP technique and resembles RRP, but it is usually applied to MPS items and a shorter horizon.

For resource r in period t:

where:

• C_r,t = estimated capacity requirement;
• M_j,t = MPS quantity of item j;
• f_j,r = average capacity hours of resource r per unit of item j;
• n = number of MPS items.

The factor may be calculated from historical data, standard costs, routing averages or a representative product mix. For example, if 500 units of Product A require 0.4 hours at a fabrication centre and 300 units of Product B require 0.7 hours, the estimated load is 410 hours.

Advantages: rapid calculation, low data requirement and suitability for early alternatives analysis.

Limitations: it ignores operation sequence, timing within the product structure, setup effects, order-specific routings and changes in product mix. It is unreliable where products have highly different routings or where capacity is concentrated in a few operations.

6. Capacity bills method

A capacity bill is a simplified representation of the resource requirements associated with producing one unit of an end item. Unlike a full manufacturing bill of material and routing, it usually records only the key capacity resources and their average hours per end item.

A capacity bill for item j may be represented as a vector:

where b_jr is the required capacity of resource r per unit of item j. The load generated by an MPS quantity is:

The calculation resembles the overall factors method, but the capacity bill can distinguish products and can be built around a selected set of bottleneck resources. It is therefore more transparent and often more accurate.

A capacity bill may include material-centre hours, assembly labour, testing hours and outside-processing hours, while omitting routine resources that are known to have ample capacity. Its quality depends on periodically updating standards when routings, technology, yields or product designs change.

7. Resource-profile method

The resource-profile method adds time-phased information to the capacity bill. It recognises that a product scheduled for completion in a particular week consumes capacity in earlier weeks according to its lead time or operation offsets.

Suppose item j has capacity requirement b_j,r,k for resource r at offset k periods before completion. Then the load in period t can be expressed as:

where:

• M_j,t+k = quantity of item j due for completion in period t+k;
• b_j,r,k = resource hours for resource r occurring k periods before completion;
• K_j = maximum profile offset for item j.

This is a time-phased convolution of the MPS with the resource profile. The formula assumes that profile requirements are standardised and that the planned offset is a reasonable approximation of actual operation timing.

Worked example: profile loading

An item has the following profile per unit:

The MPS calls for 100 units in Week 3 and 120 units in Week 4. Fabrication load in Week 2 is 20 hours, generated by Week 3 output. Fabrication load in Week 3 is 10 + 24 = 34 hours: 10 hours from Week 3 output at one week before completion and 24 hours from Week 4 output at two weeks before completion. Assembly load in Week 3 is 25 hours from Week 3 output, while Week 4 assembly load is 30 + 18 = 48 hours.

This method can reveal a Week 3 bottleneck that an unphased capacity bill would miss. It is especially useful when lead times are stable and the organisation needs an early warning before detailed order release.

8. Detailed Capacity Requirements Planning

Detailed CRP uses MRP planned orders, bills of material, routings, operation times, work-centre calendars, queue assumptions and order dates. It calculates the load on individual work centres and time periods, usually after MRP has exploded the MPS into component requirements.

For operation o of order q, the required time is commonly:

where Q_q is the order quantity, t_q,o is run time per unit, and s_q,o is setup time. If the routing specifies a yield y between 0 and 1, the quantity launched may be adjusted to Q_q ÷ y when the yield loss occurs before the operation being planned. The adjustment must reflect the actual process location of the loss.

CRP assigns operation load to work centres and periods using forward scheduling, backward scheduling or finite-capacity logic. A basic work-centre load report compares, for every period:

• available capacity;
• scheduled load;
• projected utilisation;
• overload or underload;
• orders contributing to the condition.

Detailed CRP is stronger than RCCP for execution planning because it can identify the exact order and operation causing an overload. It is also more sensitive to inaccurate data. A routing with an obsolete standard, an incorrect queue time or a missing setup can produce a precise but misleading result.

9. Comparing the methods

The methods should not be treated as competing systems in which only one is selected permanently. They are complementary layers. A robust planning process uses RRP to shape feasible aggregate plans, RCCP to validate the MPS, and CRP to control detailed execution.

10. Interpreting overloads and selecting responses

An overload is a condition in which planned requirement exceeds available capacity in a resource-period combination. It is a signal for investigation, not an automatic instruction to add permanent capacity. Appropriate responses include:

1. Reschedule: move orders to periods with spare capacity.
2. Change the MPS: alter product mix, quantity or completion dates.
3. Use overtime or additional shifts: appropriate for temporary peaks if labour and safety conditions permit.
4. Subcontract: transfer work externally after evaluating quality, cost, intellectual property and supplier reliability.
5. Alternative routing: use another capable work centre or process.
6. Reduce setup time: apply SMED, campaign planning or improved tooling.
7. Improve yield and reliability: remove quality losses, downtime and rework.
8. Add capacity: acquire equipment, recruit or redesign the process when the shortage is structural.

The economic comparison should include more than direct overtime cost. Expediting, late-delivery penalties, inventory, quality risk, employee fatigue, supplier dependence and lost sales may materially change the decision.

11. Illustrative case: Northstar Pumps

The following case is fictional and intended for instructional use.

Background

Northstar Pumps manufactures industrial centrifugal pumps. Its MPS showed rising demand for two families: Standard pumps and High-pressure pumps. Management proposed adding a weekend shift, but the operations manager requested a capacity analysis before committing.

Problem

The aggregate plan indicated adequate total labour. However, high-pressure pumps used a specialised pressure-test station and a skilled balancing cell. These resources were not visible in the aggregate labour average.

Analysis

RRP, using average labour-hours, showed 87% utilisation for the quarter and recommended no major action. RCCP was then performed using capacity bills for the pressure-test station and balancing cell. The MPS produced the following weekly loads:

RCCP identified Week 2 and Week 3 pressure-test overloads and significant balancing-cell risk. A resource profile showed that some Week 3 completions consumed balancing time in Week 2, making the near-term peak worse than the unphased monthly view suggested.

Detailed CRP traced the overload to three high-pressure orders and showed that setup time had been omitted from one routing. After correcting the routing, the load increased by 6 hours. The team evaluated alternatives: weekend overtime, subcontracted testing, a revised sequence and a one-week MPS shift.

Approach and outcome

Northstar shifted two Standard-pump orders by one week, scheduled a controlled Saturday test shift, and trained a second operator for the balancing cell. No permanent equipment was purchased. The revised plan reduced peak pressure-test utilisation to 98% and balancing utilisation to 96%, while preserving customer due dates.

Learning points

The case demonstrates that RRP can give a reassuring average while RCCP reveals a critical resource shortage. Resource profiles improve timing visibility, and CRP is necessary to investigate the precise routing and order causes. The final solution combined schedule change, temporary capacity and workforce flexibility rather than relying on a single intervention.

12. Data quality, governance and technology

Capacity planning is a socio-technical process, not only a calculation. Important master data include work-centre calendars, available resource counts, routing sequences, setup standards, run times, yields, queue times, alternative resources and maintenance schedules. Governance should define data ownership, approval rules, revision dates and exception thresholds.

Enterprise resource planning and advanced planning systems can automate capacity calculations, scenario comparison and exception reporting. Digital twins and machine-data platforms may improve estimates of actual processing time and availability. AI and machine learning can support anomaly detection, predictive maintenance and probabilistic capacity forecasts, but they should not replace process ownership. A model trained on historical schedules may reproduce past inefficiencies or fail when product mix, labour skills or technology changes.

Good practice includes:

• maintaining one controlled source for routing and calendar data;
• separating standard capacity from temporary overtime;
• testing sensitivity to yield, absenteeism and demand variation;
• reviewing bottlenecks with supervisors and operators;
• using exception-based reports rather than overwhelming planners with detail;
• preserving an audit trail for schedule changes;
• protecting employee and supplier data in digital systems.

Ethically, capacity decisions affect workload, fatigue, job security, supplier dependence and delivery commitments. Overtime should comply with labour law and occupational-health requirements. Outsourcing decisions should consider responsible sourcing, confidentiality and quality assurance, not cost alone.

13. Critical evaluation and method selection

No capacity method is universally superior. The appropriate method depends on the decision, horizon, volatility and data maturity.

Use RRP when the question concerns broad, long-term resource requirements and the product mix is uncertain or still aggregated. Use RCCP when an MPS exists and management needs rapid validation of bottlenecks before detailed release. Use detailed CRP when orders, routings and calendars are sufficiently reliable and the organisation must make near-term sequencing or release decisions.

When demand and routings are highly volatile, a deterministic calculation may be supplemented by scenarios or probability ranges. For example, planners can test a base case, a high-demand case and a reduced-availability case. The objective is not to predict one exact future but to identify decisions that remain robust across plausible conditions.

A mature organisation therefore asks four questions:

1. What decision is being made?
2. What is the shortest useful planning horizon?
3. Which resource constraints actually determine feasibility?
4. Is the data quality sufficient to justify the chosen level of detail?

Chapter Summary

Hierarchical capacity planning links strategic, tactical and operational decisions. Resource Requirements Planning uses aggregate production information and broad resource factors to assess long-range feasibility. Rough-Cut Capacity Planning tests the master production schedule against selected critical resources. Detailed Capacity Requirements Planning uses MRP orders, routings and work-centre calendars to identify operation-level loads.

The overall factors method is fast but highly aggregated. Capacity bills provide resource consumption by product and key resource. Resource profiles add time-phased offsets and are especially useful where production activities occur several periods before final completion. Detailed CRP offers the greatest diagnostic precision but requires accurate and maintained data.

Capacity requirements should include both run and setup time where applicable, and all calculations must use consistent units. Overloads may be addressed through rescheduling, overtime, alternate routing, subcontracting, process improvement or permanent investment. Effective capacity planning combines quantitative analysis with managerial judgement, data governance, workforce considerations and continuous feedback.

Key Terms

Resource Requirements Planning: Aggregate capacity planning based on the production plan and broad resource categories.

Rough-Cut Capacity Planning: Medium-range capacity validation of the master production schedule.

Capacity Requirements Planning: Detailed calculation of work-centre and operation loads from MRP orders and routings.

Aggregate production plan: A medium- or long-range plan stated in product families or output quantities.

Master production schedule: A time-phased plan for specific end items.

Load: Capacity demanded by planned or released work.

Available capacity: Time that a resource can realistically provide in a planning period.

Effective capacity: Usable capacity after planned losses and availability adjustments.

Utilisation: Required load expressed as a proportion or percentage of available capacity.

Overall factors method: RCCP method using average resource factors per MPS item.

Capacity bill: Simplified resource requirement statement for an end item.

Resource profile: Time-phased capacity requirements associated with an item’s completion.

Work centre: A machine group, labour group or facility treated as a capacity resource.

Setup time: Time required to prepare a resource for a production run.

Run time: Processing time that varies with quantity produced.

Bottleneck: A resource whose constrained capacity limits system throughput.

Finite-capacity scheduling: Scheduling that does not assign more work than available capacity.

Master data: Controlled information such as routings, calendars, standards and work-centre definitions.

Capacity overload: A condition in which planned requirement exceeds available capacity.

Scenario analysis: Evaluation of capacity under alternative assumptions or plans.

Review Questions

Short Answer Questions

1. What is the principal purpose of hierarchical capacity planning?
2. Distinguish nominal capacity from effective capacity.
3. Why is RCCP performed before detailed CRP?
4. What information does a resource profile add to a capacity bill?
5. State two limitations of the overall factors method.

Long Answer Questions

1. Compare RRP, RCCP and detailed CRP in terms of purpose, horizon, data and managerial decisions.
2. Explain the overall factors, capacity bills and resource-profile methods, including their assumptions and limitations.
3. Discuss how setup time, yield and alternative routings affect capacity calculations.
4. Explain why a capacity-planning result can be mathematically precise but operationally unreliable.
5. Evaluate the advantages and risks of using ERP, advanced planning systems and AI in capacity planning.

Analytical / Application Questions

1. A work centre has 2 machines, 2 shifts per day, 7.5 scheduled hours per shift and 20 working days. If the effective availability factor is 0.85, calculate effective monthly capacity and explain how a 520-hour load should be interpreted.
2. An MPS contains 400 units of Product A requiring 0.30 hours at Resource X and 250 units of Product B requiring 0.55 hours. If Resource X has 250 available hours, calculate the load and utilisation, then recommend two possible responses.
3. A planner finds no overload in monthly RCCP but a resource-profile report shows a two-week peak above capacity. Explain why the results differ and identify the data or scheduling decisions that should be investigated.

Multiple Choice Questions with Answers

Question 1

Which planning method primarily evaluates broad resource needs from an aggregate production plan?

A. Resource Requirements Planning

B. Detailed Capacity Requirements Planning

C. Dispatch-list control

D. Finite sequencing

Correct Answer: A. Resource Requirements Planning

Question 2

A work centre has 600 hours of effective capacity and a planned load of 540 hours. Its utilisation is:

A. 60%

B. 90%

C. 111.1%

D. 1,140 hours

Correct Answer: B. 90%

Question 3

What is the main additional feature of a resource profile compared with a simple capacity bill?

A. It eliminates setup time

B. It uses only aggregate product families

C. It time-phases resource usage relative to completion

D. It replaces the master production schedule

Correct Answer: C. It time-phases resource usage relative to completion

Question 4

Which input is most characteristic of detailed CRP?

A. Product-family forecast only

B. MRP planned orders and detailed routings

C. Annual sales budget only

D. Supplier financial statements

Correct Answer: B. MRP planned orders and detailed routings

Question 5

If a resource has 3 machines, 2 shifts per day, 8 hours per shift and 20 working days, nominal capacity is:

A. 160 machine-hours

B. 320 machine-hours

C. 480 machine-hours

D. 960 machine-hours

Correct Answer: C. 480 machine-hours

Question 6

A capacity bill normally records:

A. Only selling prices

B. Key resource requirements per unit of an end item

C. Employee performance ratings

D. Customer payment terms

Correct Answer: B. Key resource requirements per unit of an end item

Question 7

Which action is most appropriate when a temporary bottleneck is caused by a short demand peak?

A. Immediately purchase a new factory

B. Ignore the overload

C. Evaluate overtime, rescheduling or subcontracting

D. Delete the routing

Correct Answer: C. Evaluate overtime, rescheduling or subcontracting

Question 8

The overall factors method is particularly vulnerable to errors caused by:

A. Product-mix differences hidden by averages

B. Using any time-phased data

C. Including setup time

D. Maintaining work-centre calendars

Correct Answer: A. Product-mix differences hidden by averages

Question 9

If 500 units require 0.4 hours each at a resource and 300 units require 0.7 hours each, total estimated load is:

A. 200 hours

B. 410 hours

C. 560 hours

D. 800 hours

Correct Answer: B. 410 hours

Question 10

Why should an overload report not automatically lead to permanent capacity investment?

A. Overloads are always calculation errors

B. The condition may be temporary or solvable through scheduling and flexibility

C. Capacity investment never affects output

D. RRP cannot identify any resource

Correct Answer: B. The condition may be temporary or solvable through scheduling and flexibility