(B) - NETWORK PLANNING Notes
(B) - NETWORK PLANNING Notes
NETWORK PLANNING
Contents
1. Introduction ........................................................................................... 1
2. Network Representation ........................................................................ 1
3. Critical Path Networks .......................................................................... 2
4. Computer Assisted Documentation........................................................ 3
5. Strategy of Adjusting the Project Network ............................................ 6
6. Applications of Networks to Project Control......................................... 6
7. The Control Cycle.................................................................................. 7
8. Reporting ............................................................................................... 7
9. Re-scheduling for Timing ...................................................................... 8
10. Re-scheduling for Resources ................................................................. 9
NETWORK ANALYSIS
1. INTRODUCTION
Network Analysis is a generic term for a family of related techniques developed to aid
management in the planning and control of projects. These techniques show the inter
relationship of the various jobs or tasks which make up the overall project and clearly
identify the critical parts of the project. They can provide planning and control
information on the time, cost and resource aspects of the project.
(b) Large, i.e. they contain many tasks with varying resource types and costs
(c) Where restrictions exist, i.e. target completion dates, limited resources or costs
limits
2. NETWORK REPRESENTATION
The method of representing the tasks as boxes linked by lines to show their sequence
is known as the ACTIVITY ON NODE technique. The resulting networks are known
as precedence diagrams. A START node and FIMSH node are required to complete
the network.
Generally, the lengths of the lines have no significance and they represent logical
dependencies only. Usually, when one activity has finished the next in line is
assumed to start immediately. However, more accurate modelling may sometimes
require a lag time to be introduced.
3.2 Float
The difference in time between the late start and early start for a particular activity is
called the Total Float. If this is positive (i.e. the late start date is after the early start),
then that activity can be delayed by that period of time without affecting the end date
of the project.
Conversely, if the total float is negative, then the project end date will be late by the
extent of the negative float. (Note that in this case that the early time is later then the
late time, which at first seems anomalous but is explained by the above definitions
and is the cause of the float being negative).
When early and late times are the same, the float is zero. A path with zero float is
called a Critical Path. This is of academic interest because in real networks more that
one path can have zero float. Activities on late paths (with negative float) should be
manipulated (controlled) so that float is adjusted to zero. This indicates what needs to
be done so that the project can be completed by the target date.
3.3 Milestones
Intermediate activities can be set target dates. This is normal if those particular
activities are to be used as milestones. Float is then calculated against the milestone
dates although it would be over written if subsequent milestones or project end dates
resulted in lower values of float. This is a useful way of distributing float throughout
a project at the planning stage.
Float can also be affected by target start dates, i.e. NOT to be started before…...dates.
However the real power of the Critical Path Method (CPM) is that once the logic of
the project activities has been defined for arithmetic analysis then computers can be
used not only for the time analysis calculations butt also to assist with resource
scheduling, calculation of spend profiles and bar charts.
- It can access both input data and calculated data (e.g. Early Start times)
- flexibility in which data fields are present (e.g. Early Start &/or Late Start)
- flexibility in which data records are selected (e.g. activities for the whole project
or just a single Work Package)
- ease of updating documents print results to computer file for further processing
If there is sufficient float for over-demands to be avoided, then some other steps must
be taken by the project manager. The choices are:
- Review the logic or work sequence to check that tasks are not being forced in
parallel unnecessarily
When the date is plotted an S-shaped curve generally results. This may have little use
at the Launch phase of a project but becomes the base-line document during the
control cycle phase, when actual achievement and predicted performance are
superimposed as the project progresses.
Accountants require to deal with figures rather than graphs and charts; their principle
method of reporting is by variance analysis, which involves subtracting actual spend
from budget spend. When costing labour, they may separate overheads from wages,
and often want to superimpose different inflation rates for future years. The cost of
borrowing to finance a project is a function of the cash flow profile, which is the
difference between the spend profile and the payment profile. The spend profile must
correctly identify when sub-contractors are likely to submit their invoices and
appropriate times for clearing and payments, i.e., payment is not strictly tied to
completion of work.
However, although figures must be made available for accountants, it is usually the
project manager who must take decisions for controlling the work. The correct
graphical presentation can greatly ease decision making, rather than needing to
comprehend tables of numbers which usually only represent the situation at one point
in time.
Next there is a check that resources are available and to what extent peaks can be
levelled.
Finally, the cost is calculated. If this was outside the permissible budget then changes
to the starting assumptions might be necessary.
These measurements are accumulated and compared to the appropriate base line
figures. This can be a clerical task. The variance data is fed to management for
analysis and consideration of response.
Except in the case of the very short term or situations where there is an inevitable,
straightforward sequence of events, it would be rare that plans work out exactly as
originally conceived. At the start of the plan there is rarely sufficient information to
plan accurately. There is always considerable uncertainty about events in the future.
When changes occur, the new information should be fed into the plan to determine the
best decisions and resource allocations to achieve the project objectives. Thus the
first loop on the control cycle is completed and new instructions are issued for
implementation during the next cycle.
When external changes are imposed (e.g. change of specification), then the plan
should be used to evaluate the effect. Without a plan the changing circumstances can
only be approximated.
8. REPORTING
As illustrated above, the plan should be a dynamic model of the project. However, it
is to be used as a management tool it must as far as practical, be always up to date.
The input/output measurements should be taken regularly so that variance data reports
are as current as practical. If reports are delayed then managers may take decisions
on a judgmental basis without waiting for planning analysis.
However, the input figures (resources and costs so far) will generally be made
available by the company infrastructure, which will probably be on a monthly basis.
It may be difficult, if not impossible to acquire these figures more regularly.
Revised estimates to complete are usually unpopular and involve a lot of people (the
work package holders) in a lot of re-calculation. This must be endorsed by
management at all levels if meaningful figures are to result. The total effort can be
reduced by only demanding frequent re-estimates for those work packages with
adverse variances.
Planning in a high level of detail is not a solution to obtaining tighter project control.
The more detailed a plan attempts to be, the more likely it is that some aspects will
change so the plan becomes inaccurate and then confidence in it is undermined, e.g., a
plan detailing activities as short as one day is likely to need updating after one week
and the number of activities to be progressed will be more than when the same work
is programmed on a coarser scale. Thus the workload can become onerous.
The extent of the negative float indicates the extent to which the path must be
shortened. Each tack should be considered to see if a saving on duration time can be
justified (or on leads and lags if they are used). Considerations should include affects
of overtime, additional resources, chasing suppliers for earlier delivery dates or re-
sequencing of activities.
If all of these fail, then the probability is that the project will finish late. The most
late path may be adjusted to finish on time but the remaining late paths must similarly
be adjusted. Generally these paths can be treated independently and this greatly
simplifies the re-scheduling of all tasks to finish on time.
The algorithm that schedules the activities usually assumes that first priority on
resources should be given to activities with least float. It would then slip less critical
activities to times when resources were available. Clearly, the computer must be
instructed as to what the available resource profiles are.
The outcome of such software is that SCHEDULED start and finish dates are
suggested, instead of using ES, EF, LS and LF. Remaining Float is usually quoted.
overheads, i.e., project engineers' can control man hours but have little control over
overheads.
However, these figures indicate the extent of spend but not the extent of progress. An
estimate of the Value of Work Done (normally referred to as the Budget Cost of the
Work Performed - BCWP) should be made for comparison with the recorded spend so
far. However, the VOWD is often difficult to establish. If the planning has been
done in a high level of detail, then the VOWD can be calculated from the sum of the
resources originally budgeted on the activities which are now finished.
On the other hand, with a less detailed plan the individual activities will be longer so
will be more likely to be ongoing (rather than finished) so an estimate must be made
of what has been achieved.