SAP Work centers

A work center models a production resource: it defines available capacity (for capacity planning), standard formulas and times (for scheduling), and links to cost centers/activity types (for costing). Routings assign operations to work centers, so the work center is where the "where" and "how much capacity" of production live.

• Capacity: available hours/shifts.
• Scheduling data: formulas using routing times.
• Costing: cost center and activity types/rates.
• Assignment to operations via routings.

CR01 creates a work centre, CR02 changes it and CR03 displays it. CR05 lists which routings use one, and running it before any change is the habit that prevents surprises.

The screens divide into four areas worth knowing separately. Basic data holds the usage, which controls where the work centre may be used, and the standard value key, which decides which times a routing may enter against it. Capacities hold available hours, shifts, and a utilisation factor. Scheduling holds the formulas converting standard values into duration. Costing holds the cost centre and the activity types that price the time.

That last screen is the join to controlling, and it is the one most often left incomplete. A work centre with no cost centre assignment will schedule perfectly and cost nothing, which means every product made through it is undercosted and nothing warns you.

The tables are CRHD for the header and CRCA for capacity.

Half an hour, and it shows the difference between a schedule and a plan anyone can meet.

1. Display a work centre with CR03 and note its available capacity: hours per shift, shifts per day, utilisation.
2. Create several production orders that all need that work centre in the same week.
3. Look at the capacity requirements against available capacity. If scheduling is infinite, the orders were all scheduled regardless of whether the hours exist.
4. Compare the total required hours against available. Overload is visible here and nowhere on the orders themselves.
5. Level the load, either by moving an order or by adding a shift, and watch the requirement redistribute.

The point is that infinite scheduling is the default in many configurations, and it produces plans that look fine and cannot be executed. Capacity evaluation is where that shows up. See PP scheduling for how the dates are calculated.

Work centers connect several PP functions: they drive scheduling (times), capacity planning (available capacity), and product costing (activity rates from the linked cost center). Accurate work-center data is therefore essential for realistic schedules and correct manufacturing costs.

The chain is worth stating explicitly, because it explains why a work centre error is expensive. The routing supplies standard values. The work centre's formulas turn them into duration, which sets the schedule. The work centre's activity rates turn them into money, which sets the product cost. One master record feeding both the plan and the cost is unusual, and it means an error is wrong twice. See PP routing for the other half of that pairing.

• How granular. One work centre per machine gives precise scheduling and a lot of master data. One per line or cell is coarser and easier to maintain, and it is right where the line is scheduled as a unit.
• Finite or infinite scheduling. Finite respects capacity and produces realistic dates. Infinite is simpler and produces optimistic ones.
• Which activity types. Machine and labour separately, or one combined rate. Separating them gives better cost analysis and requires both to be maintained and confirmed.
• Utilisation factor. The honest allowance for breaks, changeovers and downtime. Setting it to one hundred per cent guarantees the schedule is wrong.

Having capacity data is one thing. Using it is capacity planning, and it is where PP stops being paperwork and starts changing what the plant does.

Capacity requirements are generated by every planned and production order that uses a work centre. They accumulate by day or by week, and they are compared against available capacity.

Capacity evaluation is the report showing that comparison, and it is the screen a planner works from. Overload appears as a percentage above one hundred, and the useful view is by work centre over the next few weeks rather than a single day.

Capacity levelling is the act of resolving it: move an order earlier or later, split it across periods, send it to an alternative work centre, or add a shift. Some of this can be automated with a levelling profile; much of it is a planner making judgements about which customer waits.

The prerequisite is honest data. A work centre claiming twenty-four hour availability on a plant that runs one shift produces evaluations showing everything is fine while the floor is overwhelmed. The utilisation factor is where reality is admitted: breaks, changeovers, maintenance and the fact that nobody works at full rate all day.

Finite scheduling takes this further by refusing to schedule beyond available capacity in the first place, so the plan is executable rather than optimistic. It is harder to configure and it produces dates the business can actually commit to, which is usually worth the effort where capacity genuinely constrains output. See the bill of materials for the other master record every order reads.

Plants rarely have exactly one place to do a job, and modelling the alternatives is what makes a schedule resilient.

A work centre hierarchy groups related work centres so capacity can be evaluated at a level above the individual machine. That answers the question a planner actually asks, which is whether the area can absorb the work rather than whether one machine can.

Alternative work centres can be named on an operation, so that if the primary is unavailable the work can be moved without editing the routing. Where a plant has genuinely interchangeable machines, this is the difference between rescheduling and rebuilding orders.

Production versions take it further by pairing a different routing with the same material, which is the right model when the alternative method differs in more than which machine runs it.

Downtime is represented by changing available capacity for the period, either through a capacity interval or by removing a shift. Doing that rather than leaving the capacity in place is what makes the evaluation reflect the plant, and it is frequently skipped because it takes a few minutes when everyone is busy dealing with the breakdown.

• Wrong capacity data, bad capacity planning.
• Missing costing link, wrong product costs.
• Inaccurate scheduling formulas/times.

• Costing screen left blank. The most common single error, and it fails silently by producing costs that are simply too low.
• Capacity copied from a machine that was replaced. Nobody revisits it, and the schedule quietly drifts from reality.
• Changing a shared work centre without the where-used list. See production orders for what consumes it.

• Formulas left as delivered without checking what they charge per. A setup time charged per unit instead of per order produces costs that scale absurdly with order size.

Reading a work centre is straightforward, and setting capacity, formulas and rates so both the schedule and the cost come out right is the part you do in the course.

The check worth running on any work centre you inherit is whether the costing screen is complete. It is the most common single gap, it fails silently, and every product routed through that work centre is undercosted until somebody notices.