Unit 9: Resolving Overallocations

INT416 — Software Project Management Laboratory 11 min read

I. Orientation — Resource Overallocations and Schedule Balance

A resource overallocation occurs when a resource is assigned more work than its available capacity during a particular time period. In project scheduling software such as Microsoft Project, resolving an overallocation means changing task timing, assignment load, resource capacity, or work distribution while preserving project objectives as far as possible.

Governing principle: Demand for a resource during any time period should not exceed that resource’s available capacity.
- A resource available at 100% normally represents one full-time working capacity.
- Two simultaneous 100% assignments create a demand of 200%, producing an overallocation.
Basic scheduling relationship: Work, duration, and assignment units are related approximately by:

TEXT

Work = Duration × Assignment Units

Work: Total effort, usually measured in hours.
Duration: Scheduled working time between task start and finish.
Assignment Units: Percentage or decimal representation of resource capacity.
- Resolution choices: Overallocations can be corrected by delaying or splitting tasks, reducing work, replacing resources, increasing availability, scheduling overtime, or changing how work is distributed.
- Scheduling constraint: A correction should be checked against dependencies, deadlines, fixed dates, task calendars, resource calendars, and the critical path.
- Key distinction: An overallocation is time-phased; a resource may have a reasonable total workload but still be overloaded on a specific day or hour.
- Baseline discipline: Before major corrections, preserve an approved baseline so that revised start dates, finish dates, duration, work, and cost can be compared with the original plan.

II. Schedule-Based Resolution — Changing When Work Occurs

Schedule-based methods retain the assigned resources and much of the original work but move or interrupt tasks so that competing assignments no longer occur at the same time.

A. Delaying Tasks Manually

Manual delay resolves an overallocation by moving a selected task or assignment to a later time after considering its scheduling importance.

Task delay: Change the task’s start date, create or modify a predecessor relationship, or enter task-level delay where appropriate.
Assignment delay: In assignment-oriented views, delay only one resource’s work without necessarily moving every assignment on the task.
Selection rule: Delay the task with lower priority, greater slack, a later deadline, or fewer dependent successors.
Dependency preference: A logical relationship is generally more maintainable than entering an arbitrary fixed date.
- For example, changing two parallel tasks into a Finish-to-Start sequence removes simultaneous demand.
Risk: Manual movement may extend the project finish date or create a constraint that prevents later automatic rescheduling.
Verification: Recheck the Resource Usage view and critical path after every delay because shifting one task may create another conflict.

B. Leveling Resources

Resource leveling is the automatic process of delaying or splitting assignments until resource demand fits the available capacity.

Operation: Microsoft Project examines assignments, calendars, dependencies, constraints, priority, and available slack before selecting tasks to delay.
Typical access: Use the resource-leveling commands on the Resource tab, choose leveling options, and then level all resources or a selected resource.
Calculation timing:
- Automatic: Leveling occurs when the schedule changes.
- Manual: Leveling occurs only when the user issues a leveling command, allowing greater control.
Granularity: The “look for overallocations on a…” setting may use minute-by-minute, hourly, daily, weekly, or monthly periods.
Important limitation: Leveling normally changes timing; it does not independently reduce task work, change assignment units, or add staff.
Outcome check: Compare the leveled completion date and total slack with the pre-leveling schedule.

C. Leveling Resources Continued

Advanced leveling options control which tasks may move and how the leveling algorithm treats priorities and task structure.

Leveling order:
1. ID Only: Gives strong influence to task identification order.
2. Standard: Considers dependencies, slack, dates, constraints, and priority.
3. Priority, Standard: Uses task priority first, followed by standard scheduling factors.
Priority values: Task priority ranges from 0 to 1000; a value of 1000 generally prevents leveling from delaying or splitting that task.
Available slack option: “Level only within available slack” avoids extending the project finish but may leave some overallocations unresolved.
Adjustment scope: Leveling may be permitted to adjust individual assignments on a task rather than moving the entire task.
Splitting permission: If splitting is allowed, the leveler can interrupt remaining work and resume it when the resource becomes available.
Clear leveling: Clearing leveling removes leveling delays and splits, but it does not reverse unrelated manual schedule changes.

D. Using The Leveling Gantt View

The Leveling Gantt view displays the schedule before and after leveling so that imposed delays can be inspected visually.

Visual comparison: Distinct bars show the original or preleveled position and the current leveled task schedule.
Delay evidence: A visible gap between the preleveled and scheduled bars identifies how far a task has moved.
Table information: Fields such as leveling delay, start, finish, and task priority explain why the bar changed.
Diagnostic use: Trace delayed tasks through their successors to determine whether leveling altered a milestone or project completion.
Control use: If an unacceptable task moved, undo or clear leveling, adjust its priority or logic, and run leveling again.
Interpretive caution: A visually delayed task is not necessarily harmful; delay within free or total slack may leave successor and project dates unchanged.

III. Work and Task-Pattern Resolution — Changing the Demand Profile

These techniques modify the quantity or continuity of scheduled effort. They require managerial judgment because they may affect scope, productivity, quality, or task execution.

A. Decreasing Work Time

Decreasing work reduces the effort demanded from an overloaded resource and is valid only when the task can genuinely be completed with fewer labor hours.

Legitimate bases: Remove unnecessary scope, automate an activity, simplify a deliverable, reuse an existing component, or revise an inflated estimate.
Direct edit: Reduce the Work field for the task or assignment after obtaining agreement from the responsible manager.
Task-type effect:
- On a fixed-units task, less work generally produces shorter duration.
- On a fixed-duration task, less work generally reduces assignment units.
Example: If work falls from 40 hours to 32 hours at 100% units, the calculated duration can fall from five eight-hour days to four.
Quality safeguard: Work must not be deleted merely to make the warning symbol disappear; required deliverables and acceptance criteria remain controlling.
Cost effect: Reduced regular work usually lowers labor cost, subject to fixed costs and rate tables.

B. Splitting Tasks

Splitting divides a task into separate working segments, creating a nonworking interruption between them.

Purpose: Pause lower-priority work while an overloaded resource completes a more urgent assignment, then resume the interrupted task.
Procedure: In a Gantt view, use the split-task command and position the resumed segment at the required date.
Scheduling behavior: The task retains one identity, but its duration spans working segments plus the intervening gap.
Suitable work: Documentation, coding, analysis, or other activities that can stop and restart with manageable disruption.
Unsuitable work: Continuous processes such as concrete curing, live data migration, or equipment calibration may not tolerate interruption.
Cost and productivity risk: Restarting may introduce setup time, context switching, or repeated verification not represented in the original estimate.
Dependency check: Successors normally wait for the final segment to finish, so a split can consume slack or delay later work.

IV. Slack-Based Diagnosis — Finding Safe Scheduling Flexibility

Slack, also called float, measures how far a task can move without violating a scheduling boundary. Viewing slack before changing dates helps distinguish safe delays from changes that threaten milestones or project completion.

A. Viewing Slack In A Table View

A table view provides numerical slack values that support precise selection of tasks for delay or leveling.

Total slack: Time a task can be delayed without delaying the project finish date or a controlling deadline.
Free slack: Time a task can be delayed without delaying the early start of any successor.
Display method: Apply a schedule-oriented table or insert the Total Slack and Free Slack fields into a task sheet.
Interpretation:
- Positive slack indicates scheduling flexibility.
- Zero slack commonly identifies a critical task.
- Negative slack indicates that the current schedule conflicts with a constraint, deadline, or required finish.
Decision use: When two simultaneous tasks overload one resource, the task with sufficient positive slack is normally the safer candidate for delay.
Calendar basis: Slack is expressed in elapsed scheduling units interpreted through the applicable project, task, and resource calendars.

B. Viewing Slack In A Gantt View

A Gantt view represents slack graphically, making the relationship between task flexibility, successors, and the critical path easier to see.

Display method: Use a Detail Gantt or another configured Gantt view that includes slack bars and critical-task formatting.
Graphical meaning: A slack bar extending beyond the task bar shows the interval available before the controlling date is affected.
Critical-path comparison: Critical tasks typically have zero or minimal total slack and are commonly emphasized with a distinct bar color.
Resolution use: Move or level noncritical bars within their displayed slack before altering critical work.
Combined analysis: Use graphical slack to recognize patterns, then consult table fields for exact durations such as “2 days” or “8 hours.”
Limitation: Gantt scaling can conceal short conflicts; an hourly overallocation may be invisible when the timescale is displayed by month.

V. Capacity and Assignment Resolution — Changing Who Works and When

Capacity-based methods preserve task timing by transferring work, expanding availability, using overtime, or redistributing effort across an assignment.

A. Reassigning Resources

Reassignment transfers some or all work from an overloaded resource to another suitable and available resource.

Replacement criteria: Match skills, authorization, experience, location, equipment access, availability, and cost rate.
Full replacement: Remove the overloaded resource and assign a qualified substitute to the task.
Partial reassignment: Divide assignment work between resources when the task is divisible and coordination costs are acceptable.
Schedule effect: Adding a resource does not always reduce duration; task type, effort-driven scheduling, and the nature of the work determine the result.
Cost check: Compare standard rates, material rates, and any onboarding or handover effort before confirming the change.
Practical constraint: Specialized work may have no interchangeable resource, making delay, overtime, or scope adjustment more realistic.

B. Increasing Working Hours

Increasing working hours expands a resource’s calendar availability during selected periods without necessarily changing task work.

Calendar method: Modify working time in the resource calendar by extending a shift or making a normally nonworking period available.
Example: Changing availability from eight to ten hours for one day provides two additional schedulable hours on that date.
Temporary use: Enter exceptions with defined start and finish dates rather than permanently changing the base calendar.
Distinction from units: Increasing maximum units adds capacity through availability percentage; extending the calendar adds working time.
Human constraint: Longer schedules must comply with organizational policy, labor law, fatigue limits, and health and safety requirements.
Schedule check: Calendar changes can pull work earlier or shorten duration, depending on task type and assignment settings.

C. Assigning Overtime To A Task

Overtime assigns part of existing work to hours outside the resource’s normal working schedule and applies the resource’s overtime rate.

Assignment field: Enter Overtime Work for the resource assignment, usually through a task or resource usage view.
Work relationship: Overtime work is a portion of total assignment work, not additional scope automatically added to the task.