Detailed CRP Process and Time-Phased Load Calculation

Introduction

Capacity Requirements Planning (CRP) translates the output of material requirements planning (MRP) into a time-phased statement of work-centre requirements. MRP answers primarily what materials or components are needed and when; CRP asks whether the available labour, machines, tools and supporting resources can perform the scheduled operations in the required periods. CRP is therefore a capacity-feasibility and reconciliation process rather than a simple report-generation exercise.

The central managerial problem is dynamic. A work centre may have adequate total capacity over a month but still be overloaded in a particular week or day. Conversely, apparently sufficient capacity may be unusable because it is located in the wrong work centre, occurs in the wrong time bucket, or depends on unavailable skills, tooling or maintenance windows. Effective CRP must consequently preserve operation sequence, lead-time offsets, order status and the distinction between theoretical and usable capacity.

This chapter develops a complete CRP process: preparing input data, converting planned and released orders into operation loads, offsetting operations, constructing load profiles, interpreting overloads and underloads, and closing the feedback loop by reconciling capacity decisions with the master production schedule and MRP.

Learning Objectives

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

1. Explain the relationship among the master production schedule, MRP, CRP and shop-floor control.
2. Identify the data required for reliable work-centre load calculation.
3. Convert planned and released orders into operation-level capacity requirements.
4. Offset operations through routing lead times and distinguish scheduling conventions.
5. Calculate gross, available and net capacity by work centre and time bucket.
6. Construct and interpret load profiles and capacity reports.
7. Diagnose overloads and underloads using quantitative and managerial criteria.
8. Apply a closed-loop process for revising schedules, routings, capacity or priorities.
9. Evaluate the limitations, governance risks and digital developments associated with CRP.

1. CRP within the Production Planning Hierarchy

A typical planning hierarchy contains several linked decisions:

• The sales and operations plan establishes aggregate production, inventory and workforce intentions.
• The master production schedule (MPS) specifies end-item quantities and completion periods.
• MRP explodes end-item requirements through bills of material and produces planned orders or released orders for components and assemblies.
• CRP uses those orders and their routings to calculate work-centre loads.
• Shop-floor control dispatches, monitors and reports actual operations.

CRP is neither independent of MRP nor a substitute for finite scheduling. Traditional CRP often assumes that planned order dates are fixed and then compares calculated load with available capacity. Finite-capacity scheduling goes further by sequencing jobs while explicitly preventing resource use above a defined limit. In practice, CRP is frequently used as a medium-term feasibility check, while detailed scheduling resolves short-term sequencing and dispatching.

The feedback loop is essential. If a work centre is overloaded, the planner may change order timing, lot sizes, routing, overtime, subcontracting, staffing, tooling, the MPS or even the aggregate plan. A capacity report that merely identifies overloads without initiating corrective action is not a complete CRP system.

2. Data Architecture and Planning Assumptions

2.1 Required master data

Reliable CRP depends on the following information:

A routing is more than a list of departments. It should identify operation numbers, work centres, setup requirements, run-time standards, overlap rules, queue or move times, and possible alternate resources. Standard times should be periodically validated against observed performance, while changes should be controlled so that historical and future plans remain interpretable.

2.2 Orders used in CRP

CRP may include:

• Planned orders, which are MRP recommendations not yet converted into shop orders.
• Firm planned orders, which are protected from automatic rescheduling within a defined time fence.
• Released production orders, which have been authorized for execution.
• Open operations, representing work already in process.
• Actual recorded work, which should replace, not be added to, the corresponding standard requirement where the system design requires it.

The inclusion policy must be explicit. Including both an original released order and a successor rescheduled order, for example, creates duplicate load. Similarly, completed operations must not remain in the future-load calculation.

2.3 Time buckets and horizons

CRP commonly uses daily buckets for short-term control and weekly buckets for medium-term planning. A bucket is a reporting interval, not necessarily a scheduling rule. Aggregating a five-day load into a weekly total can conceal a daily overload. The planning horizon should cover the longest relevant cumulative lead time plus enough visibility for capacity decisions such as hiring, subcontracting or equipment acquisition.

3. Operation-Level Load Calculation

For an operation, total required time normally consists of setup time and run time. If setup is performed once per order or batch, the requirement is:

where L_i,w,t is the load in hours for order or operation i assigned to work centre w in time bucket t, S_i,w is setup time in hours, Q_i is the quantity processed, and R_i,w is run time per unit in hours per unit. The formula assumes one setup for the relevant batch and that all units follow the same standard.

If an order is split across batches, setup is incurred for each batch:

Here B_i is the number of batches. This distinction is important because smaller batches may reduce flow time but increase total setup load.

If labour and machine time differ, they should be calculated separately rather than combined:

A single operation can therefore consume one hour of machine time and two hours of labour time, especially where an operator supervises multiple machines or performs preparation activities. Capacity comparisons must use the resource type being constrained.

For partial completion, the remaining load should be calculated from the remaining quantity or remaining standard work. If Q^rem units remain and setup has already been completed, the remaining run load is:

If setup has not begun, setup must be added. Actual progress should be reconciled with system status to avoid counting completed work again.

Worked Example 1: Operation load

A released order requires 240 units. Its routing contains three operations:

The loads are:

• Cutting: 1.5 + (240 × 0.08) = 20.7 hours.
• Milling: 2.0 + (240 × 0.12) = 30.8 hours.
• Inspection: 0.5 + (240 × 0.03) = 7.7 hours.

The total is 59.2 hours across three different work centres. It is incorrect to compare 59.2 hours with the capacity of any single work centre.

4. Time-Phasing and Operation Offsetting

Load calculation determines how much work is required; time phasing determines when that work is expected to occur. The required completion date of the final operation is traced backward through the routing. In a simplified backward schedule, operation start dates are obtained by subtracting operation duration and relevant inter-operation times from the required completion date.

For a serial routing with operation duration d_k and inter-operation offset m_k, where k identifies an operation, a backward relationship can be written as:

Here T^start_k and T^finish_k are the start and finish times, d_k is operation duration, and m_k-1,k is the required move, wait or transfer time between operations. Units must be consistent, such as working hours or working days.

In many MRP systems, routing lead times are represented by fixed operation offsets rather than calculated from actual hourly capacity. A planner must therefore know whether the system uses:

• Backward scheduling, placing work as late as possible while meeting the due date;
• Forward scheduling, starting at an available release date and projecting completion; or
• Overlap scheduling, allowing the next operation to begin before the preceding operation finishes, subject to a transfer quantity or overlap rule.

Calendar time and working time must not be confused. A two-hour operation scheduled across a lunch closure or a non-working day does not consume two calendar hours in the same way as two continuous hours on the shop calendar. The work-centre calendar should determine valid start and finish points.

Worked Example 2: Backward time phasing

An order of 100 units is due at the end of Day 10. The routing is:

Assume one working day contains 8 hours and that the due point is the end of Day 10.

Operation 30 must finish at the due point. It starts 3 hours earlier, or 0.375 day, during Day 10. Operation 20 must finish 0.5 day before operation 30 starts; it therefore finishes at 0.125 day into Day 10 and starts 0.75 day earlier, at 0.375 day into Day 9. Operation 10 must finish one day before operation 20 starts, at 0.375 day into Day 8, and starts 0.5 day earlier, at the end of Day 7.

The example illustrates why load cannot be assigned solely to the due-date bucket. Work centres A, B and C receive load in different periods. In a real system, exact dates would be rounded to shop-calendar intervals and adjusted for holidays, breaks, queue time and available capacity.

5. Available, Demonstrated and Net Capacity

Nominal capacity is the scheduled time before losses. For work centre w in bucket t, a basic expression is:

where N_w,t is the number of available machines or operators and H_w,t is scheduled working hours per resource.

Usable or demonstrated capacity should account for planned losses and operating performance. A common planning approximation is:

where A is availability, P is performance relative to standard, and Y is the proportion of output meeting the relevant quality criterion. These factors should be used carefully. If standard time already incorporates observed performance or if quality losses do not consume the constrained resource, multiplying by all three factors double counts losses.

A simpler approach deducts known time losses directly:

The correct method depends on how standards and calendars are maintained. The essential requirement is conceptual consistency.

Net load and utilization are commonly reported as:

Here L^gross is total required load, C^committed is capacity already reserved for valid work or protected commitments, and U is utilization. If U > 1 or 100%, the work centre is overloaded under the stated assumptions. A negative net load indicates nominal spare capacity, not necessarily immediately usable capacity.

Worked Example 3: Capacity comparison

A work centre has two machines, each scheduled for 8 hours per day over five days. Planned maintenance is 4 machine-hours, and the planning policy uses an 85% effective factor for the remaining scheduled time.

Nominal weekly capacity is:

2 × 8 × 5 = 80 machine-hours.

After maintenance, usable scheduled time is 76 hours. Applying the factor gives:

76 × 0.85 = 64.6 machine-hours.

If the CRP load is 70 hours, utilization relative to usable capacity is:

70 ÷ 64.6 ≈ 1.084, or 108.4%.

The overload is 5.4 hours. However, the planner must confirm that the 85% factor does not already include maintenance; otherwise the calculation understates capacity.

6. Building Load Profiles and Capacity Reports

A load profile displays required hours and available hours by work centre and time bucket. A basic report should show at least:

Variance may be defined as L - C, so a positive value indicates overload. Some organizations use the reverse sign; the report must state its convention.

A useful profile distinguishes:

1. planned-order load;
2. released-order load;
3. open work-in-process load;
4. firm or protected load; and
5. alternate-resource or subcontractable load.

This separation supports managerial decisions. A projected overload caused only by tentative planned orders is treated differently from an overload caused by released orders already committed to customers.

Reports should permit drill-down from a work-centre total to period, order, operation, product, setup and routing version. Aggregate percentages alone are insufficient for corrective action. Visual displays may use bars for load and a line for capacity, but the underlying data should remain auditable.

7. Diagnosing Overloads and Underloads

An overload is not automatically a reason to add permanent capacity. It may arise from inaccurate standards, duplicated orders, excessive safety lead time, an incorrect calendar, poor lot sizing, a temporary demand peak or an invalid routing. Diagnosis should proceed from data integrity to timing, then to operational remedies.

7.1 Typical overload responses

• Reschedule: Move planned orders to earlier or later buckets, subject to material availability and customer commitments.
• Change lot size: Split a batch, combine setups, or revise order quantities when economically and technically appropriate.
• Use alternate resources: Transfer work to qualified machines or work centres.
• Add time: Authorize overtime, extra shifts or weekend work after checking labour and safety constraints.
• Subcontract: Outsource selected operations while considering quality, intellectual property and supplier capacity.
• Improve the process: Reduce setup, remove bottlenecks, improve yield or revise the routing.
• Revise the MPS: Change end-item timing or quantity when the constraint cannot be economically removed.
• Acquire capacity: Hire, lease or purchase equipment when the pattern is persistent and strategically justified.

7.2 Underloads and false slack

Underload may indicate excess capacity, but it may also reflect missing orders, late component availability, excessive planned downtime, a bottleneck elsewhere, or a skill constraint that prevents the nominal resource from being used. Capacity is usable only when material, labour, tooling, quality approval and supporting resources are available simultaneously.

A work centre with 20 spare machine-hours cannot necessarily absorb 20 hours of work if the required operator skill is unavailable. Similarly, shifting all work to an alternate machine may create a downstream inspection constraint. CRP decisions should therefore consider the network of linked constraints rather than isolated percentages.

8. Closed-Loop CRP Process

A disciplined closed-loop process consists of the following cycle:

1. Establish the planning version. Freeze or identify the MPS, MRP run, calendars, routings and standard-time revision used in the analysis.
2. Select valid demand and orders. Include planned, firm, released and open orders according to policy; remove completed or superseded records.
3. Explode routings. Convert each order into operation-level requirements, including setup, run, labour, machine and tooling demands.
4. Offset operations. Place each operation in time using the stated scheduling convention and shop calendars.
5. Aggregate load. Sum load by work centre and bucket, preserving resource type and order status.
6. Calculate capacity. Apply calendars, planned downtime, staffing, efficiency assumptions and alternate-resource rules.
7. Compare and prioritize. Identify overloads, underloads, critical periods and protected commitments.
8. Investigate causes. Validate master data, material availability, order duplication, standards and timing.
9. Apply corrective actions. Evaluate schedule changes, overtime, subcontracting, routing alternatives, process improvement or MPS revision.
10. Re-run MRP and CRP. A capacity action can alter dates, component requirements and downstream loads; therefore one pass is rarely sufficient.
11. Authorize and monitor. Record decisions, owners and due dates, then compare planned load with actual performance.

This loop should be governed by time fences. Near-term released work may be protected, while distant planned orders may be moved more freely. Exception-based management is generally preferable: planners focus on material overloads, constraint violations and changes beyond tolerance rather than manually reviewing every line.

Control metrics

Useful indicators include forecast or schedule volatility, percentage of overloaded buckets, load accuracy, schedule adherence, capacity utilization, overtime hours, queue time and the proportion of orders requiring manual intervention. A high average utilization can coexist with poor delivery performance if variability and queues are high. Capacity reports should therefore be interpreted alongside flow and service measures.

9. Detailed Illustrative Case: Meridian Pumps

The following is a fictional illustrative case.

Background

Meridian Pumps manufactures industrial pump assemblies. Its MPS requires 120 assemblies in Week 4 and 160 in Week 5. MRP generates component and assembly orders. The impeller line contains machining, balancing and inspection work centres. Management has recently observed late deliveries despite a reported monthly utilization below 90%.

Problem

The initial CRP report aggregates load by month. It shows adequate capacity, but customer orders due in Week 5 are frequently late. The operations manager requests a weekly, operation-level analysis.

Analysis

The relevant routing for one pump is:

The 160-unit Week 5 order is processed in two batches of 80. Its machining load is:

2 × 3 + 160 × 0.40 = 70 hours.

Balancing load is:

2 × 1 + 160 × 0.15 = 26 hours.

Inspection load is:

2 × 0.5 + 160 × 0.10 = 17 hours.

The routing requires machining to occur before balancing and inspection. Backward offsetting places 70 machining hours partly in Week 4, while the initial monthly report assigned the entire requirement to Week 5. The weekly work-centre profile shows machining at 112% of usable capacity in Week 4 and 98% in Week 5; balancing and inspection remain below 85%.

A data audit identifies two additional issues. First, the machining calendar included planned maintenance as available time. Second, the standard setup was recorded once per order even though the production policy required two batches.

Approach and solution

The team corrects the calendar and setup rule, then evaluates alternatives. The preferred response is to move one batch of 80 units earlier, perform the two setups during a planned overtime window, and use a qualified alternate machine for 16 machining hours. The MPS is not changed because customer dates can be met without reducing output. The revised CRP is rerun after MRP regeneration to ensure component dates remain feasible.

Outcome

The corrected profile exposes a genuine Week 4 constraint rather than a monthly capacity problem. The plan meets the Week 5 requirement, and the organization avoids purchasing equipment based on misleading monthly utilization. Actual results are monitored: setup time, machining hours, queue time and completion dates are compared with standards.

Learning points

1. Aggregate capacity can conceal bucket-specific overloads.
2. Routing and batch policy materially affect load.
3. Capacity depends on valid calendars, not only machine count.
4. A local remedy must be checked for downstream and material consequences.
5. CRP is a recurring feedback process, not a one-time calculation.

10. Technology, Governance and Contemporary Practice

Enterprise resource planning systems automate much of the explosion, offsetting and aggregation, but automation does not guarantee correct results. Master-data governance should define ownership of routings, standard times, calendars, alternate resources and order statuses. Every CRP run should have a traceable planning version and audit trail.

Advanced planning and scheduling systems increasingly combine CRP with finite-capacity scheduling, constraint programming, digital twins and real-time shop-floor data. Machine-learning models may improve estimates of processing time, downtime or yield, but they should not silently replace approved standards. Predictions need validation, explainability appropriate to the decision, monitoring for drift and clear accountability. A model that systematically underestimates load can create unsafe overtime and unreliable commitments.

Integration with manufacturing execution systems enables actual start, completion, scrap and downtime information to update remaining load. Internet-connected equipment can improve visibility, but cybersecurity, access control, data quality and business continuity remain governance requirements. Capacity decisions also have ethical implications: overtime and schedule compression must comply with labour law, fatigue controls and safety requirements. Subcontracting should respect quality, confidentiality, labour and environmental obligations.

11. Limitations and Best Practices

CRP is an approximation when standards, calendars or lead times are inaccurate. It may also assume infinite intermediate storage, stable processing times or unlimited material availability. Traditional bucketed CRP can hide sequence-dependent setups, queueing, blocking and starvation. For highly variable environments, simulation, finite scheduling or queuing analysis may be needed in addition to CRP.

Best practices include:

• maintain controlled, periodically validated routings and standard times;
• separate machine, labour, tooling and skill loads;
• use working calendars and explicit time-zone or shift conventions;
• distinguish planned, firm, released and actual work;
• report both absolute hours and utilization percentages;
• drill down from exceptions to orders and operations;
• use realistic demonstrated capacity without double counting losses;
• protect near-term commitments through time fences;
• test alternate resources and downstream effects;
• rerun MRP and CRP after material schedule changes;
• compare plan with actuals and update standards through controlled procedures; and
• document assumptions, decisions, approvals and unresolved risks.

Chapter Summary

Capacity Requirements Planning converts MRP orders into time-phased work-centre requirements. The process begins with valid orders, routings, calendars, standard times and resource definitions. Operation load is commonly calculated as setup time plus quantity multiplied by run time, with multiple setups recognized when batches are split. Loads must then be offset through the routing so that each operation appears in the period in which it is expected to occur.

Capacity comparisons require a clear distinction between nominal and usable capacity. Utilization, variance and net load are useful indicators, but they do not replace diagnosis of data quality, material availability, skills, tooling, sequencing and downstream constraints. Overloads may be addressed through rescheduling, lot-size changes, alternate resources, overtime, subcontracting, process improvement, MPS revision or long-term investment. Underloads may represent genuine slack or hidden constraints.

The most effective CRP practice is closed-loop: calculate, compare, investigate, act, regenerate MRP, rerun CRP and monitor actual performance. Digital systems can improve speed and visibility, but governance, auditability, human judgment, safety and data quality remain essential.

Key Terms

Capacity Requirements Planning (CRP): A process that translates planned or released orders into time-phased resource requirements.

Work centre: A machine, group of machines, labour pool or other resource where operations are performed.

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

Operation: A defined manufacturing or inspection activity in a routing.

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

Run time: Processing time associated with each unit or quantity produced.

Planned order: An MRP-generated recommendation not yet authorized for execution.

Released order: A production order authorized for shop-floor execution.

Time bucket: A planning interval such as a day or week.

Time phasing: Assigning requirements to the periods in which they are expected to occur.

Backward scheduling: Scheduling operations backward from a required completion date.

Nominal capacity: Scheduled resource time before relevant losses or effectiveness adjustments.

Usable capacity: Capacity available after calendars, downtime and approved effectiveness assumptions.

Gross load: Total calculated requirement before subtracting capacity or other offsets.

Net load: Load remaining after an applicable capacity or committed-capacity deduction.

Utilization: The ratio of required load to usable capacity.

Overload: Load exceeding the capacity available in a specified resource and period.

Underload: Capacity exceeding calculated load in a specified resource and period.

Time fence: A planning boundary that limits automatic changes near-term.

Closed-loop planning: Repeated planning, capacity comparison, corrective action and replanning.

Review Questions

Short Answer Questions

1. What is the principal difference between MRP and CRP?
2. Why must setup time be treated differently when an order is split into batches?
3. What is the purpose of backward scheduling in CRP?
4. Distinguish nominal capacity from usable capacity.
5. Why can a monthly capacity report conceal an operational problem?

Long Answer Questions

1. Explain the complete closed-loop CRP process, including the role of MPS, MRP and shop-floor control.
2. Discuss the master data required for reliable CRP and the consequences of poor data governance.
3. Explain how planned, firm and released orders should be treated differently in a capacity report.
4. Compare rescheduling, overtime, alternate resources, subcontracting and capital investment as responses to overload.
5. Critically evaluate the limitations of bucketed CRP in environments with sequence-dependent setups and high variability.

Analytical / Application Questions

1. A work centre has three machines scheduled for 7.5 hours per day over five days. Planned maintenance consumes 6 machine-hours, and an effectiveness factor of 0.90 applies to the remaining time. A weekly load of 105 hours is expected. Calculate usable capacity and utilization, then recommend two questions the planner should investigate before authorizing overtime.
2. An order of 300 units is divided into three batches. A routing operation has a setup time of 2.5 hours per batch and a run time of 0.18 hours per unit. Calculate total operation load and explain how the result would change if the order were processed as one batch.
3. A weekly profile shows a machining overload in Week 3, an inspection underload in Week 3 and adequate monthly capacity overall. Develop a closed-loop action plan that tests whether the issue is caused by timing, routing, setup policy, calendar data or a genuine bottleneck.

Multiple Choice Questions with Answers

Question 1

Which statement best describes the primary purpose of CRP?

A. To calculate finished-goods selling prices

B. To translate MRP orders into time-phased work-centre loads

C. To replace the bill of material with a routing

D. To record financial transactions from production

Correct Answer: B. To translate MRP orders into time-phased work-centre loads

Question 2

An operation requires a 2-hour setup and 0.25 hours per unit for 100 units processed in one batch. What is its load?

A. 25 hours

B. 27 hours

C. 52 hours

D. 102 hours

Correct Answer: B. 27 hours

Question 3

If a 200-unit order is processed in four batches, which expression correctly represents its load when setup is 1.5 hours per batch and run time is 0.10 hours per unit?

A. 1.5 + 200 × 0.10

B. 4 × 1.5 + 200 × 0.10

C. 4 × (1.5 + 0.10) × 200

D. 200 × 1.5 + 4 × 0.10

Correct Answer: B. 4 × 1.5 + 200 × 0.10

Question 4

What is the main purpose of backward scheduling?

A. To schedule all operations in the due-date bucket

B. To place work as late as possible while meeting the required completion date

C. To ignore operation sequence

D. To maximize setup frequency

Correct Answer: B. To place work as late as possible while meeting the required completion date

Question 5

A work centre has 80 hours of usable capacity and 92 hours of load. What is its utilization?

A. 86.96%

B. 92.00%

C. 115.00%

D. 172.00%

Correct Answer: C. 115.00%

Question 6

Which item is most likely to create duplicate CRP load?

A. Removing a completed operation

B. Using a valid work-centre calendar

C. Including both a superseded order and its replacement

D. Separating labour and machine capacity

Correct Answer: C. Including both a superseded order and its replacement

Question 7

Which response is most appropriate for a temporary, verified overload on a qualified work centre when customer dates cannot move?

A. Delete the overload from the report

B. Consider authorized overtime or an alternate qualified resource

C. Permanently purchase equipment without analysis

D. Ignore the overload because monthly capacity is adequate

Correct Answer: B. Consider authorized overtime or an alternate qualified resource

Question 8

Why should machine and labour loads sometimes be calculated separately?

A. They always have identical capacities

B. An operation can consume different amounts of machine and labour time

C. Labour is never a capacity constraint

D. Machine time is not part of CRP

Correct Answer: B. An operation can consume different amounts of machine and labour time

Question 9

Which capacity expression is conceptually appropriate when nominal capacity is adjusted by availability, performance and yield factors?

A. Nominal capacity ÷ all three factors

B. Nominal capacity × availability × performance × yield

C. Nominal capacity + all three factors

D. Load × availability × performance × yield

Correct Answer: B. Nominal capacity × availability × performance × yield

Question 10

Why must CRP normally be rerun after a major schedule or capacity action?

A. Because capacity actions can alter dates, component requirements and downstream loads

B. Because routings become unnecessary after the first run

C. Because released orders are automatically completed

D. Because utilization must always equal 100%

Correct Answer: A. Because capacity actions can alter dates, component requirements and downstream loads