Capacity Requirements Planning and Schedule Feasibility

Introduction

Material Requirements Planning (MRP) converts the master production schedule and product structure into time-phased requirements for components and materials. However, a material plan is not automatically a feasible production plan. It may call for more work than a department, machine, supplier, or labour group can perform during the required period. Capacity Requirements Planning (CRP) provides the essential link between material planning and operational reality.

CRP translates planned and released orders into workload at work centres. It compares required capacity with available capacity, identifies overloads and underloads, and supports corrective decisions. Schedule feasibility asks a broader question: can the proposed orders be completed on time, with the available resources, under actual operating constraints?

At postgraduate level, CRP should be understood not merely as a calculation procedure but as a managerial control system. Its quality depends on the accuracy of routings, processing times, calendars, inventory records, order priorities, and assumptions about variability. A mathematically balanced plan can still fail if data are obsolete or if capacity is treated as infinitely flexible.

Learning Objectives

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

1. Explain the relationship between MRP, CRP, shop-floor control, and finite-capacity scheduling.
2. Translate planned orders into work-centre loads using routing data.
3. Calculate available capacity, required capacity, utilization, and capacity load profiles.
4. Distinguish infinite-capacity planning from finite-capacity scheduling.
5. Diagnose bottlenecks and evaluate alternatives for resolving overloads.
6. Assess schedule feasibility under lead-time, sequence, labour, and machine constraints.
7. Apply CRP principles to a realistic production-management case.
8. Critically evaluate the limitations of capacity models and the role of digital technologies.

1. From Material Plans to Work-Centre Loads

MRP determines what is required and when it is required. CRP determines where and whether the work can be performed. The basic information flow is:

Master production schedule → MRP planned orders → order quantities and due dates → routings and operation times → work-centre loads → capacity analysis → revised schedules or resources.

A routing specifies the sequence of operations required to manufacture an item. For each operation it may identify:

• work centre or machine group;
• setup time;
• run or processing time per unit;
• queue, move, and inspection time;
• labour skill requirements;
• tooling and fixture requirements;
• alternative resources;
• operation precedence;
• yield, scrap, or rework assumptions.

For a planned order, the load at a work centre is derived by multiplying the order quantity by the time required per unit and adding setup time. If several orders use the same work centre during a period, their loads are aggregated.

For order j at work centre w, required capacity is commonly represented as:

where R_wj is required capacity in hours, S_wj is setup time in hours, Q_j is order quantity in units, and t_wj is processing time per unit in hours. This expression assumes that setup is performed once per order and that processing time is constant for all units. If multiple setups, batches, or product variants exist, the setup term must be adjusted accordingly.

Total load for work centre w in period t is:

The summation is valid when all loads are expressed in the same time unit and assigned to the correct period according to the routing and lead-time logic.

Example 1: Calculating a Work-Centre Load

A work centre receives three orders in one week:

The required load is:

• A: 2.0 + (120 × 0.08) = 11.6 h
• B: 1.5 + (80 × 0.10) = 9.5 h
• C: 1.0 + (50 × 0.12) = 7.0 h

Therefore:

The work centre requires 28.1 hours before considering additional allowances for downtime, absenteeism, or quality losses.

2. Available Capacity and Capacity Measures

Available capacity is the productive time that a work centre can realistically provide during a planning period. It is not necessarily equal to the number of clock hours. Capacity must reflect shifts, breaks, planned maintenance, holidays, staffing, efficiency, and expected availability.

A simple gross-available-capacity model is:

where C_g is gross capacity in hours per planning period, N is the number of identical resources, H is scheduled hours per resource per day, D is the number of operating days, and S is the number of shifts per day when H is stated per shift. Units must be consistent.

A more realistic effective-capacity expression is:

where A is availability, P is performance efficiency, and Y is the yield factor. These factors are proportions between 0 and 1. Availability reflects time lost to breakdowns and other stoppages; performance reflects operating speed relative to a standard; yield reflects the proportion of output meeting requirements. This multiplicative model is useful for planning, but it assumes the factors are sufficiently independent. If they are correlated, historical empirical capacity should be preferred.

Required capacity should also include relevant allowances. If a basic load is R_b and an allowance factor k covers expected losses, then:

Here k is normally greater than or equal to 1. It should not be used to conceal poor standards or inaccurate routings. Losses should be separately analysed whenever possible.

Important measures include:

A load ratio above 100% indicates an overload under the stated assumptions. A ratio below 100% does not guarantee feasibility because sequence-dependent setups, material shortages, precedence relationships, and due dates may still cause failure.

Example 2: Effective Capacity

A machine group has two machines, each scheduled for 8 hours per day over five days. Gross capacity is:

Historical availability is 0.90, performance is 0.85, and yield is 0.98. Effective capacity is:

If the planned load is 63 hours, the effective load ratio is:

The centre has an approximate 3-hour deficit relative to effective capacity, even though its gross schedule appears to provide 17 hours of unused time.

3. CRP Planning Levels and Data Requirements

CRP is commonly performed at several levels:

The levels should be integrated. Rough-cut planning may use product families and average rates, whereas CRP uses individual items and routings. Finite scheduling must consider exact operation sequences, machine eligibility, release status, and real-time disruptions.

Core CRP data include:

1. MRP order records: planned orders, released orders, quantities, start dates, and due dates.
2. Routing files: operation sequence, work centre, setup time, run time, and alternatives.
3. Work-centre files: number of machines, calendars, shifts, labour availability, maintenance, and efficiency.
4. Inventory and status records: quantities completed, queued, in process, rejected, or awaiting inspection.
5. Planning parameters: lot sizes, safety capacity, overlap rules, queue assumptions, and lead-time offsets.

Data governance is particularly important. A routing that understates processing time creates apparent capacity and repeated lateness. A routing that overstates time produces unnecessary overtime, outsourcing, or investment. Standards should therefore be reviewed after engineering changes, new equipment, process improvement, or sustained variance between planned and actual performance.

4. Infinite-Capacity and Finite-Capacity Planning

Traditional CRP often uses an infinite-capacity assumption: orders are loaded into the period in which the routing indicates they should occur, even if the resulting load exceeds available capacity. This approach is useful diagnostically because it reveals where the material plan creates overloads. It does not itself produce a feasible schedule.

Finite-capacity scheduling, in contrast, refuses to schedule more work than the available resource can perform. It may move orders to later periods, select alternative machines, split lots, change sequences, authorize overtime, or trigger rescheduling upstream. Finite scheduling is therefore a constraint-based decision process rather than only a reporting process.

The distinction is important:

• Infinite-capacity CRP answers: “What load would the material plan create if timing were accepted?”
• Finite-capacity scheduling answers: “What schedule can actually be executed under stated constraints?”

A finite schedule must respect operation precedence. If operation 2 cannot begin until operation 1 is complete, moving operation 2 earlier is infeasible even if machine capacity is available. It must also respect resource eligibility, labour qualifications, tooling, material availability, and order-release rules.

5. Identifying Bottlenecks and Assessing Feasibility

A bottleneck is a resource whose capacity persistently constrains system throughput. A temporary overload is not necessarily a structural bottleneck; it may result from an unusual order mix, poor sequencing, inaccurate standards, or a short-term disruption.

Bottleneck analysis should consider:

• load ratio over several periods rather than one period only;
• queue length and waiting time;
• utilization and actual throughput;
• schedule adherence;
• dependence on scarce labour, tools, or suppliers;
• variability and recovery time after disruption;
• impact on downstream due dates.

A useful feasibility test requires more than total hours. Consider an order with operations on Centres 1 and 2. Even if both centres have sufficient weekly capacity, the order may be late if Centre 2’s available time occurs before Centre 1 completes its work. Thus, feasibility is both a capacity and a timing problem.

A practical feasibility checklist asks:

1. Are all required materials and components available at release time?
2. Is each operation assigned to an eligible resource?
3. Does every work centre have sufficient capacity in the relevant time bucket?
4. Are operation precedence and minimum transfer times satisfied?
5. Are labour, tooling, inspection, and maintenance constraints represented?
6. Can the plan absorb normal variability without immediate lateness?
7. Does the schedule meet customer due dates and priority rules?

6. Resolving Overloads

Overload resolution should follow a structured hierarchy. The best decision is not always to add capacity. Managers should first determine whether the load is valid, whether the due date is essential, and whether the bottleneck is truly constraining system output.

6.1 Data and process corrections

Before changing the schedule, validate quantities, routing versions, setup assumptions, scrap factors, calendars, and order status. Eliminate duplicated orders and correct obsolete engineering data. A false overload can be more damaging than a real one because it leads to unnecessary cost.

6.2 Routing and lot-size decisions

Routing alternatives may transfer work to an underused machine or qualified supplier. A process may also be redesigned to eliminate an operation, combine operations, or reduce setup time. Lot splitting can release completed sublots earlier, although it may increase handling and transport costs.

Reducing lot size can lower waiting time and improve responsiveness, but it may increase setup frequency. The decision must compare the reduction in flow time with the additional setup burden.

6.3 Sequencing and finite scheduling

Common sequencing rules include earliest due date, shortest processing time, critical ratio, and bottleneck-first dispatching. No rule is universally best. Earliest due date supports delivery performance, while shortest processing time often reduces average flow time. Critical ratio can be expressed as:

Both numerator and denominator must use the same time unit. A critical ratio below 1 indicates that the remaining work exceeds the remaining time under the stated assumptions. The rule is a prioritization aid, not proof that an order is feasible.

6.4 Capacity and sourcing decisions

Other responses include overtime, additional shifts, temporary labour, cross-training, subcontracting, equipment leasing, preventive maintenance changes, and capital investment. Each alternative has cost, quality, lead-time, and risk implications. Overtime may solve a short disruption but become unsustainable if used to compensate for chronic under-capacity.

A managerial response should distinguish short-term tactics from structural solutions:

7. Worked Application: Multi-Period Capacity Profile

A fabrication centre has effective capacity of 42 hours per week. The MRP plan creates the following loads:

The cumulative deficit over Weeks 2 and 3 is 11 hours. However, the surplus in Weeks 1 and 4 cannot automatically solve it. Week 1 capacity may be unusable if materials are not available or if the operations are not yet released. Week 4 capacity may occur after customer due dates.

Suppose 5 hours of Week 2 work can be moved from Week 3 to Week 4 without violating due dates, while 6 hours can be subcontracted at a premium. The revised profile is:

The remaining Week 2 deficit is still 7 hours. If 4 hours of overtime and 3 hours of cross-trained labour are available, the schedule becomes feasible, subject to labour and quality approval. This example illustrates that capacity balancing is time-specific; total monthly capacity is not sufficient evidence of schedule feasibility.

8. Illustrative Case: Orion Pumps

Background

Orion Pumps manufactures industrial centrifugal pumps. Its MRP system generated a four-week plan after a large customer order was accepted. The plan appeared feasible when reviewed against total monthly machine hours.

Problem

The machining work centre had a nominal monthly capacity of 640 hours and a planned load of 615 hours. Nevertheless, several orders were late. A weekly review showed that Week 2 required 178 effective hours against 150 available hours. The problem was concentrated in a high-precision boring operation performed by two qualified machines and a small group of certified operators.

Analysis

The initial routing file understated setup time for three product families and did not reflect a recent inspection requirement. In addition, the MRP lot-sizing rule grouped orders into large batches. This reduced the number of setups on paper but created long queues and delayed downstream assembly.

The capacity team recalculated loads using actual setup standards and effective machine availability. It also separated the boring operation from general machining, because aggregate centre-level capacity had concealed the constraint.

Approach and solution

Orion adopted four measures:

1. Corrected setup and inspection times in the routing master.
2. Split two large lots into transfer batches so that assembly could begin earlier.
3. Sequenced jobs by due date and product-family similarity, balancing delivery urgency against setup reduction.
4. Cross-trained one operator and subcontracted a limited number of non-critical boring operations.

A finite schedule was then generated for the next four weeks. The company retained a small protective capacity allowance at the constrained centre rather than planning it continuously at 100%.

Outcome

The revised plan reduced the Week 2 overload from 28 hours to 6 hours. The remaining gap was covered by approved overtime. Average queue time at boring fell, and assembly received partial lots earlier. Monthly machine utilization fell slightly, but on-time delivery improved because the actual bottleneck was managed explicitly.

Learning points

The case demonstrates that:

• aggregate monthly capacity can conceal time-bucket overloads;
• accurate routings are a prerequisite for meaningful CRP;
• setup reduction and lot splitting involve trade-offs;
• a bottleneck should be analysed at the operation level, not only at the department level;
• lower utilization can be economically desirable when it improves flow and delivery reliability.

9. Technology, Analytics, and Governance

Enterprise resource planning systems automate much of the CRP calculation, but automation does not eliminate judgment. Advanced planning and scheduling systems can use constraint programming, mixed-integer optimization, simulation, and heuristic search. Digital twins may combine machine status, labour availability, and planned orders to update capacity estimates continuously.

Artificial intelligence can support anomaly detection, predicted downtime, setup-time estimation, and schedule re-optimization. Its outputs should remain auditable. A model trained on historical schedules may reproduce undesirable practices, such as systematically delaying low-volume customers or overloading particular employee groups. Human planners must therefore review objective functions, constraints, data quality, and exception logic.

Governance issues include:

• version control for routings and calendars;
• authorization of changes to capacity standards;
• traceability of schedule overrides;
• protection of supplier and employee data;
• validation of algorithmic recommendations;
• escalation procedures when a system schedule conflicts with safety or quality requirements.

Technology is most valuable when it improves visibility and decision speed while preserving accountability. A sophisticated scheduler using inaccurate routings remains inaccurate at high speed.

10. Limitations and Critical Evaluation

CRP models simplify reality in several ways. Processing times may vary with operator, product condition, batch size, temperature, tool wear, or learning effects. Capacity may be divisible in a model but indivisible in practice. Labour cannot always be transferred instantly between centres, and machines may require specific tooling or certification.

The assumption of deterministic capacity is especially problematic in high-variability environments. A plan loaded to 100% of expected capacity has little protection against breakdowns, absenteeism, urgent orders, or rework. Conversely, excessive protective capacity may increase cost. The appropriate buffer depends on demand variability, recovery options, customer priorities, and the economic cost of lateness.

CRP also tends to treat orders as independent, whereas the factory is a network. Improving one non-bottleneck centre may have no effect on throughput if the bottleneck remains unchanged. Managers should therefore combine CRP with bottleneck analysis, queueing insight, lean flow principles, and financial evaluation.

Best-Practice Framework

A robust CRP process can be organized as follows:

1. Validate the master schedule and MRP order data.
2. Confirm routing versions, setup standards, calendars, and resource eligibility.
3. Calculate infinite-capacity loads by work centre and time bucket.
4. Compare loads with effective capacity, not merely clock hours.
5. Identify persistent bottlenecks and critical overloads.
6. Test schedule feasibility using precedence, materials, labour, tools, and due dates.
7. Evaluate alternatives using cost, delivery, quality, safety, and strategic criteria.
8. Generate a finite schedule and communicate assumptions.
9. Monitor actual performance against standards and revise the model.
10. Conduct regular feedback from shop-floor control to MRP and master scheduling.

Chapter Summary

Capacity Requirements Planning translates MRP orders into time-phased workloads at specific work centres. Required capacity depends on order quantity, setup time, processing time, routing sequence, and relevant allowances. Available capacity must reflect calendars, resource count, shifts, maintenance, availability, performance, and yield.

Infinite-capacity CRP reveals the load implied by a material plan, whereas finite-capacity scheduling creates a schedule that respects resource limits and operational precedence. Schedule feasibility requires more than an aggregate comparison of hours; it also depends on materials, timing, labour, tools, machine eligibility, sequence, and due dates.

Overloads may be addressed through data correction, routing changes, setup reduction, lot splitting, sequencing, overtime, cross-training, subcontracting, or investment. Bottlenecks should be distinguished from temporary overloads and analysed across time. Digital planning and AI can improve visibility and optimization, but they require accurate data, governance, transparency, and managerial oversight. The central managerial principle is to balance capacity, flow, delivery, cost, and resilience rather than maximize utilization in isolation.

Key Terms

Capacity Requirements Planning: A process that translates planned orders into work-centre capacity requirements.

Available Capacity: Productive resource time available during a specified planning period.

Required Capacity: Time needed to perform planned work at a work centre.

Work Centre: A machine, department, labour group, or resource pool performing specified operations.

Routing: The ordered sequence of operations required to manufacture an item.

Setup Time: Time required to prepare a resource before processing a batch or order.

Processing Time: Time required to perform an operation, often expressed per unit.

Load Ratio: Required capacity divided by available capacity, usually expressed as a percentage.

Effective Capacity: Capacity adjusted for availability, performance, yield, or other losses.

Infinite-Capacity Planning: Loading orders without preventing assignments that exceed resource capacity.

Finite-Capacity Scheduling: Scheduling that explicitly respects resource capacity constraints.

Bottleneck: A resource that constrains system throughput.

Critical Ratio: Remaining time until due date divided by remaining work.

Lot Splitting: Dividing an order into smaller transfer batches.

Rough-Cut Capacity Planning: Aggregate capacity analysis used to test the broad feasibility of the master schedule.

Shop-Floor Control: Operational monitoring and dispatching of work in process.

Capacity Buffer: Reserved capacity intended to absorb variability or disruption.

Routing Accuracy: Degree to which routing data represent actual operations and times.

Schedule Feasibility: The condition in which a plan can be executed within resource, timing, material, and precedence constraints.

Review Questions

Short Answer Questions

1. What is the primary purpose of Capacity Requirements Planning?
2. Distinguish gross capacity from effective capacity.
3. Why can monthly capacity totals fail to demonstrate schedule feasibility?
4. What information is contained in a manufacturing routing?
5. Define a bottleneck and distinguish it from a temporary overload.

Long Answer Questions

1. Explain how MRP orders are translated into work-centre loads, including the role of setup and processing times.
2. Compare infinite-capacity CRP with finite-capacity scheduling and discuss the appropriate use of each.
3. Discuss alternative methods for resolving a work-centre overload and the trade-offs associated with each.
4. Explain why accurate master data and governance are essential for reliable CRP.
5. Evaluate the benefits and limitations of applying AI and advanced planning systems to capacity scheduling.

Analytical / Application Questions

1. A work centre has three machines scheduled for 7.5 hours per day over five days. Availability is 0.92, performance is 0.88, and yield is 0.97. Calculate effective weekly capacity and determine the load ratio if required work is 86 hours.
2. A planned order requires 3 hours of setup and 0.06 hours per unit for 400 units. Calculate required capacity. If the work centre has 28 effective hours available, determine the capacity deficit and propose two feasible responses.
3. A factory has sufficient total weekly hours but repeatedly misses due dates at one operation. Develop a diagnostic approach that considers time buckets, routing accuracy, precedence, bottleneck status, sequencing, and labour constraints.

Multiple Choice Questions with Answers

Question 1

Which expression calculates the required capacity for one order when setup is performed once?

A. Q ÷ (S + t)

B. S + (Q × t)

C. S × Q × t

D. Q − S − t

Correct Answer: B. S + (Q × t)

Question 2

A work centre has required capacity of 48 hours and available capacity of 40 hours. What is its load ratio?

A. 83.3%

B. 100%

C. 120%

D. 160%

Correct Answer: C. 120%

Question 3

Which statement best describes infinite-capacity CRP?

A. It prevents all overloads automatically

B. It loads work without restricting assignments that exceed capacity

C. It sequences every job using earliest due date

D. It excludes routing information

Correct Answer: B. It loads work without restricting assignments that exceed capacity

Question 4

A machine group has gross capacity of 80 hours, availability of 0.90, performance of 0.85, and yield of 0.98. Effective capacity is approximately:

A. 59.98 hours

B. 65.20 hours

C. 72.00 hours

D. 79.20 hours

Correct Answer: A. 59.98 hours

Question 5

Which condition is necessary for a schedule to be feasible?

A. Every work centre operates at 100% utilization

B. All orders are processed in the same sequence

C. Operation precedence and resource constraints are satisfied

D. Monthly capacity exceeds monthly demand, regardless of timing

Correct Answer: C. Operation precedence and resource constraints are satisfied

Question 6

The critical ratio is calculated as:

A. Work remaining divided by available machines

B. Time remaining until due date divided by work remaining

C. Due-date quantity divided by setup time

D. Available capacity divided by demand variability

Correct Answer: B. Time remaining until due date divided by work remaining

Question 7

Which action is most appropriate before authorizing expensive additional capacity?

A. Ignore the overload until the end of the month

B. Validate routing, calendar, quantity, and order-status data

C. Increase all processing-time standards arbitrarily

D. Move every order to the next period

Correct Answer: B. Validate routing, calendar, quantity, and order-status data

Question 8

Why can lot splitting improve schedule performance?

A. It eliminates all setup time

B. It allows completed sublots to move downstream earlier

C. It guarantees lower total processing time

D. It removes the need for material availability

Correct Answer: B. It allows completed sublots to move downstream earlier

Question 9

Which measure identifies the difference between available and required capacity?

A. Capacity surplus or deficit

B. Critical ratio only

C. Scrap percentage only

D. Routing sequence index

Correct Answer: A. Capacity surplus or deficit

Question 10

Which is a significant governance concern when AI is used in scheduling?

A. AI always requires infinite capacity

B. Algorithmic recommendations may be opaque or reproduce historical bias

C. AI eliminates the need for accurate routing data

D. AI prevents all machine breakdowns

Correct Answer: B. Algorithmic recommendations may be opaque or reproduce historical bias