How to do Cost Management for Project Management?

A fool and his money are soon parted.

16th century proverb

Introduction

This section considers the aspects of capital cost (Capex) and operating cost (Opex) estimating for opportunities. In 4.2.2, levels of accuracy at various stages of the Capex process are discussed, estimating methods and allowances and contingencies are considered, consideration is given to the cost estimating requirements for areas new to
Shell and the use of stretch targets are commented on. The need for a more rigorous approach than using percentages of Capex as the Opex estimate is discussed in 4.2.3 and monetary commitment controls and approvals are discussed in 4.2.4.

Capex Estimating

This section outlines the SIEP Cost Engineering methodology, which is described in the
Cost Engineering System Manual (CESM).
The key steps in the preparation of a Capex estimate for a project are as follows:

• define the nature and key parameters or ‘scope’ of the project or development scenario being considered;
• breakdown the project into building blocks or ‘Hardware Items’, and to a level of detail appropriate for the type of estimate required;
• further breakdown each hardware item into discrete activities or ‘Project Functions’;
• derive the engineering or physical quantities of each Hardware Item using a method which takes account of key parameters of the project. For example: reservoir plateau rates for oil, gas and water, reservoir depth and area, location, terrain and distance from export market;
• estimate the cost of each project function by application of unit cost rates to the derived engineering quantities for each hardware item. For example: engineering manhours, equipment and bulk material weights, fabrication man-hours, hook-up and commissioning man-hours;
• add appropriate allowances and contingencies to the individual estimates at hardware item or project function level;
• phase the components of the total cost estimate to obtain an expenditure profile which reflects the project schedule;
• summate and record the complete estimate from the definition of scope, through the derivation of quantities and the application of unit cost rates to obtain the final estimated cost of the project.

These steps are required in all estimates. However, the reason for the estimate and the accuracy required will determine the degree of definition and the extent to which the project needs to be broken down into building blocks. The type of estimate performed should therefore be commensurate with the purpose for which it was prepared. The SIEP Cost Engineering practice is therefore:

• to select an estimate type based on the purpose of the estimate and the scope definition available;
• to adopt a predetermined breakdown structure for the selected estimate type;
• to follow a consistent estimating method across the complete breakdown structure.

This can be summarized in the following cost estimating elements which are addressed in the subsequent sections:

• estimate levels and their accuracy
• breakdown structure
• estimating method
• allowances and contingencies
• cost phasing

Estimate levels and their accuracy EP cost estimates are classified as screening, study, budget or control estimates, also referred to as Level 1 to 4, to give an indication of the accuracy which may be assigned to the cost figures. These levels of accuracy can also be related to the ORP Phases, as shown in Figure 4.2.2. The accuracy
associated with each level is a function of the level of detail of the scope definition and the variance in both the derived quantities and the unit cost rates selected for the estimate. For instance, the expected accuracy of an estimate is expressed as +X%/-Y%, for instance +25%/-15% for a study estimate, (refer Figure 4.2.2). The upper and lower boundaries of the accuracy range are defined as having a probability of less than 10% of overrun and underrun respectively. In other words an estimate with a value of 200 and an expected accuracy of +25%/-15% would have a 10% probability that the actual cost will exceed 250 and a 10% probability that it will be or be less than 170.

As will be explained in “Allowances and Contingencies” below, the accuracy range as determined by the estimator plays a role in the definition of the confidence level of the estimates.

Breakdown Structure

The selection of building blocks from which to compose the hardware required for the particular development scenario under study is the second step in the estimating process.
It is important that the boundaries of these building blocks are clearly defined, as this determines the estimating methods to be used. For all levels of estimating, the building blocks/work breakdown structure should be consistent.
The breakdown structure adopted for the SIEP cost engineering methodology is based on a hardware item/project function matrix, whereby hardware items form the physical building blocks of a project and project functions are the discrete elements of work performed on the Hardware Items (these definitions are explained in the CESM).
For Level 2 estimates, a more detailed breakdown of hardware is used to obtain more accurate estimates than for a Level 1 estimate. The breakdown structure in project functions is kept essentially the same for all estimate levels.

Estimating method

The next step in the cost estimating process involves the definition of a minimum number of
parameters, which together describe the scope of the development scenario. These parameters
for a field development, for example would include location, reservoir depth, production
plateau etc. When these have been determined, a selection can be made of hardware items
and system groups/systems in order to compose an engineering development scheme.
The hardware selected for the development under study can be defined in terms of
engineering quantities. These quantities include weight of substructure steel in tonnes,
drilling time for wells in days, design time for production facilities in man-days etc. Each
of the parameters defining the scope of the development will have an impact on one or
more of the engineering quantities. For instance, water depth will impact on substructure
weight, reservoir depth on drilling time, and so forth.
Each project function executed to realise the project, will incur a cost depending on the
quantities involved. These costs can be expressed as unit cost rates such as fabrication
cost for substructures in US$/tonne or drilling cost for wells in US$/day etc. These rates
are specific to both the Hardware Item and the Project Function in question. Cost rates
and norms should also relate to the region or country and market condition that are
expected to prevail during the execution of the project.
A cost estimate of a particular development is therefore produced by translating the
scope of the development into engineering quantities which are then multiplied by unit
cost rates per project function to arrive at cost.

In summary:

Establishing and maintaining a corporate cost engineering database is necessary for supporting norms and cost rates.

Cost considerations for new areas

Opening up areas of the world where western oil companies have not operated before
present major challenges. One of those challenges is cost estimating (both capex and
opex), especially if the country is in transition from a closed to an open economy. The
challenge will not be limited to costing the scope and operation of the facilities but will
also encompass negotiating the commercial terms with Governments or their agencies
and deciding and costing the infrastructure necessary to support the opportunity. Figure
4.2.3 provides a preliminary checklist for both negotiation with Governments and
costing the necessary work but a full listing pertinent to a particular opportunity should
be obtained by way of a framing exercise as described in 2.2 – Defining the Opportunity.
The experience of previous similar projects should be actively sought both in framing the
issues and seeking cost information.

Checklist for the interrelationships between negotiating activity and the scope and estimation of
opportunities in new areas.

Allowances and contingencies

Through the process described in the previous chapters, an estimate is obtained of
development cost associated with the particular scope defined for the estimate.
There is a tendency, however, for the scope of the unique type of projects executed by EP
to grow. As a consequence, typically greater quantities are required than estimated
in the base estimate. To cater for this growth of scope, contingencies and allowances are
added at various points in the estimating process.

Obviously, a more detailed definition of scope not only improves the accuracy of the
estimate but also reduces the levels of contingency to be applied.
Allowances and contingencies are assigned at three stages in the estimating process:
Activity allowances
Activity allowances are added by the estimator to the various cost items to account for
weight growth, weather downtime, cut and waste, and other such known uncertainties,
which have a high probability of occurring. Activity allowances are added to the various cost

items at the discretion of the estimator. The resulting estimate is named “base estimate”.
Contingency
Contingency is added to the base estimate to allow for incomplete project definition,
estimate omissions, engineering and construction market and exchange rate uncertainties
etc. (but not for major scope changes, each of which would require a new project
estimate). Current practice, in many cases, is to add what is considered reasonable at the
discretion of the estimator, to cover the unknown uncertainties. However, typical
contingencies for each estimate are shown in Figure 4.2.2 and should be used as the
default in the absence of better information. The resulting estimate is named “50/50
estimate”, refer to Figure 4.2.2.
By definition of this estimate the project has an equal chance of overrunning as of
underrunning the 50/50 estimate within its accuracy range. The contingencies are
therefore real cost elements, albeit for “unspecified” scope and are thus related mainly to
quantities. Contingency should never be used as comprehensive cover for each and every
uncertainty, or as insurance against ‘force majeure’ risks.
The “50/50 estimate” should be used as the basis for the analysis of development
economics and, at FID, as the target for project expenditure.
Overrun allowance
An allowance is added to the 50/50 estimate to allow for the risk of overrunning this
estimate. A project estimated to cost, for instance 200, with an accuracy of +25%/-15%
may require an actual expenditure of 250. Thus, the accuracy range could be taken as the
monetary “exposure” of the project. Therefore, by adding a value representative of the
accuracy range (and named: the overrun allowance) to the 50/50 estimate, an estimate is
arrived at with only 10% probability of being exceeded by the actual cost.
This estimate is named a “90/10 estimate” and could be considered a “minimum risk”
estimate. It may be used for the setting of budget levels or the sensitivity analysis of field
development economics.

Cost phasing and escalation

Once the project estimate of the required confidence level has been established, the
expenditure of cost over time can be determined. This expenditure profile should reflect
not only the durations associated with design, procurement, construction etc. of the
individual Hardware Items but also the various lead-times as dictated by the overall
project schedule. Most estimating techniques contain various methods which help to
establish an expenditure profile which reflects these considerations.
Cost estimates for Levels 1, 2 and 3 should be prepared initially in Constant Value
Money (CVM) or Real Terms (RT), referenced to a base reference date, and this should
form the basis for the cost phasing. Estimates for Level 4 and sometimes even for Level 3
are developed in Money of the Day (MOD) as actual cost rates or contract terms are
known. Care should always be taken if CVM values are generated for economic
evaluations from MOD values. The de-escalation should always match the economic
model inflation assumptions. All control estimates (Level 4) should be presented in MOD.
The phased cost estimate may then be presented in either CVM with a specified
reference date, or escalated to MOD. Escalation, which estimates the future combined
effect of general inflation and market conditions specific to the project, is necessary if the
phased cost estimate is to be used for economic analysis.

Deterministic vs. probabilistic estimates

Base, 50/50 and 90/10 estimates are terms used to specify the level of confidence to be
assigned to an estimate. Allowances and contingencies are cost elements included in these
estimates to arrive at certain levels of confidence.
Level 1 and 2 estimates;
howevercheck the Level 3 estimates.
Strictly speaking, these terms refer to two different methods of estimating; that is the
conventional (also referred to as deterministic) method and the probabilistic method.
The intention of the latter is to produce a probability curve on the basis of a detailed
probabilistic investigation of a conventionally derived base estimate.
A few years ago, probabilistic cost estimating (or cost risk analysis), began to receive
increasing attention and was expected to rapidly replace conventional methods
of estimating. In anticipation of this, the concept of confidence level and associated
terminology such as 50/50 and 90/10 estimates were introduced. However, a few
practical but fundamental problems have hindered the advancement of the technique,
particularly in ORP, Phase 1, Identify and Assess, and Phase 2, Select, because of the
impossibility of modelling the uncertainties present at that time. However, probabilistic
estimates can be successfully produced at the definition stage (ORP Phase 3) as
uncertainties can by then, be realistically modelled and correlated. Also, the process of
developing the probabilistic estimate is valuable in providing insights into the details of
the cost and schedule and this is as important as the end result. Meanwhile, terms like
50/50 and 90/10 have become household words in EP, and provide a common
understanding of the cost basis. More conventional terms such as “most likely” or
“minimum risk” probably introduce confusion, and are therefore best avoided.
(It is recognised that in the absence of reliable cost estimating probabilistics, the purpose
of the definition given to confidence level indicators or the boundaries of the accuracy
range is somewhat limited. It does, however, provide a basis for consistent use of
estimating terminology and definitions and thus for the future validation of current
practice with regard to accuracy and contingency.)
In order to validate currently applied contingencies and expected accuracy ranges, a
comparison is required of original development plans and associated budget estimates
with as-built facilities and cost data for a significant number of projects. This in turn will
require the collection of data on past and current projects in terms of scope and cost.
The CES contains cost reporting methods for this purpose. This data can be fed into the
CES which will assist in determining if the correct values for accuracy and contingency
are being applied. Metric benchmarking by independent companies can also assist in
assessing contingency and accuracy levels. (See 4.5.6)

Preparation of Level 3 (budget) estimate

The project definition at the end of Front End Engineering is at a stage that will allow
the development of a Level 3 (budget) estimate, with an accuracy of around +15%/-10%.
The Level 3 estimate is used in the economic analysis to determine whether the project is
economically viable and subsequently to seek funds for project execution.
The estimate should be developed using suitable OU or contractor developed Level 3 cost
estimating methodologies. The scope definition in the project specification permits, for
example, equipment lists to be costed using budget vendor quotes for all tagged
equipment and all bulk material take offs can also be priced, e.g. piping broken down by
size and specification, cabling similarly broken down. All major influences on the estimate
should be identified and incorporated in the cost estimate.

The estimate should be based on the actual and forecast market activity that will affect
the project over its execution phase. Special care should be taken during periods of high
market activity when the project will be competing with others for limited resources.
However, project teams should not be slow to recognise the potential savings (lower
prices) which prevail during low market activity.
The preparation activity of the budget request should take account of third-party issues.
For example care should be taken on projects with co-venturers. Reference should be
made to Joint Operating Agreements (JOA) and decisions taken as to how the estimate is
presented and what level of funding is requested. It may be prudent to consider for a
”fast track” multi co-venturer development the approval of a 90/10 budget, thereby
limiting the risk of delay if the project exceeded the 50/50 estimate.
All budget estimates for large projects should be supported by a probabilistic risk analysis.

Stretch Targets

It is now common practice to establish stretch targets on cost and schedule as a challenge
to the team to perform better than the 50/50 estimates.
In order for stretch targets to remain a stimulating challenge and not become a
demoralising burden, there needs to be a clear distinction between the team’s “contract”
with management (i.e. the 50/50 estimate) and the stretch targets. The “contract” should
be based on comparable benchmarking (i.e. outcomes achieved by others either externally
or internally) whereas stretch targets should be established that will challenge the team to
do things in more effective ways than previously.

Opex estimating

Formerly, economic models relied on a simple ratio of Capex to define the associated life cycle
operations costs. 2% – 8% was by no means uncommon. This gave two distinct problems:
• There was no real accuracy and comparison between Capex/Opex in concepts.
• There was a discontinuity in the life cycle at start up with no real base line with
which to measure operation performance.
For Levels 1 and 2 cost estimating, CES now incorporates algorithms based on
percentage Capex (dependant on the hardware items) and throughout that are sufficient
at this level for economic evaluation. However, at ORP Phase 3 – Define, activity based
costing of Asset Reference Plans (ARPs).
should be employed. This provides a similarly
rigorous methodology for Opex estimating as for Capex.
The methodologies includes:

1. Personnel requirements
   • identify owner and contractor personnel required per operation request activity per asset (onshore and offshore)
2. Resource tariff rates developed for
   • Personnel (onshore and offshore)
     both company and contractor
   • Equipment and materials
     subsurface/surface facilities:
     movables/consumables: decision support/telecom/ICT
   • Maintenance/operations
   • Services
     subsurface/surface facilities: transport/utilities/supply base
   • Product export/transport
   • Deferment
   • BOE x No. days x production rate
3. Cost activities
   • collate 1 and 2 into an activity cost model, and categorise personnel/materials/
     services/deferment.
4. Cost of operations per annum
   • Cost of operations per annum
5. Planning assumptions
   • document planning assumptions in terms of personnel/materials/services/deferment.

What is meant by Deferment in Project Management?

The deferment assumptions usually represent double cost penalties. These include:
• safeguarding technical integrity
• preventative and corrective maintenance
• production operations
• surplus capacity
• field development
• technical and other support

The more accurate the Opex model the better. Note, however, that it’s key use at the
concept selection stage (ORP Phase 2) is for comparative purposes, and hence absolute
accuracy is not required. However, there is the increasing need to improve the accuracy
of Opex estimating in order not to affect project economics adversely and increasing
emphasis is being given to its accuracy at FID. By the time detailed design is nearing
completion, the updated ARP should be providing a detailed, costed Operations plan
against which actual performance can be measured over the first two years of operation.

Commitment Control and Approvals

Introduction
No matter how large or complex a project might appear, the same four levels of work
commitment and expenditure control are required. (Figure 4.2.3) These levels permit the
delegation of responsibility and its corresponding accountability from Company/
shareholder level to the supervisor directly in charge of the work component (the work pack).

Figure 4.2.4, outlines
the relationships
between Budget,
Technical, and
Commitment
authority. Budgets are
approved subject to at
least high level
technical authorization
being in place, and
often the budget and
external technical
authorities are
interdependent and
pursued in parallel.
Commitment
authorization follows

and is the last check prior to committing funds, especially to third-parties via contracts
and procurement agreements. The budget and commitment procedures are structured
as a check and balance to avoid risk of hastily made commitments turning sour through
immature planning and staff work.
Approval of the budget by shareholders and of the project execution plan by Company
management empowers the engineer to go ahead with full (or partial) project
implementation, obtaining ‘step-by-step’ authorization from Requests For Authority

(RFAs)
11
. These are the equivalent of opening individual bank accounts (for each RFA),
against which commitments can be made and funds spent, providing the conditions of
technical and commercial approvals are met.

Commitment Control

Purpose
For a project to be delivered within cost, there needs to be a system in place which allows
the project manager to demonstrate to management (the asset holder) that planned
commitments will be within the projects approved budget. It also provides a means of
early warning that cost overrun might be possible. Management for individual
commitments, or groups of commitments, are made through RFAs as and when required.
The RFAs relating to the commitments required for the project execution phase are set
out in the commitment plan. This is part of the project execution plan and is based on
the control estimate: accurate to +10%/-5%. The total amount of the commitment plan
must always be within the approved budget.
Projects by definition require the making of commitments. In the wider sense these are
of two types:

1 External
are formal commitments in the form of contracts and purchase orders to thirparties; generally referred to as ‘commitments’ in the contractual/ legal sense
2 Internal
are agreements with in-house departments for the provision of work and/ or
materials (e.g. ex stock)

To make a commitment, the engineer must:

• follow Company contracting procedures
• follow Company authorisation procedures
• have a satisfactory commitment and expenditure reporting system

Commitment Plan

The Commitment Plan is a document which gives a financial breakdown, at RFA
level of the project over time. It shows the RFAs required to implement the project,
and the time span to which they apply. Each RFA covers a group of one or more
commitments. Each commitment represents a package of one or more activities, in
accordance with the work breakdown structure dictated by the contracting strategy.
The RFA values are based on the control estimate figures for the commitment groups
and activity packages they encompass, but with the addition of contingency to bring
the total amount up to the project budget level.
The purpose of the commitment plan is to highlight the current state and structure of
the project budget, to allow management a final review of the approved project budget
immediately before commitments are made, and to allow monitoring of commitments
to provide earliest warning of slippage. The commitment plan also includes work and/
or supplies to be provided by in-house departments. It should be submitted for approval
prior to the first RFA.
The main commitment plan is often preceded at the development plan stage by an initial
commitment plan. This not only serves as a forerunner, but should also accompany
RFAs for any commitments which have to be made before the main commitment plan
is ready; for instance, if the project specification is to be prepared by a contractor.

Preparation and approval

Approval of the commitment plan is obtained from Company management on the basis
of the proposed breakdown of the project into RFAs, representing groups of commitments.
The total amount of the commitment plan must always be within the approved budget.
A commitment is made up of one or more activities, for each of which a CTR sheet
must be raised.
The breakdown must be hierarchical along the lines of Figure 4.2.1, the cost breakdown
triangle. These not only become the basis for commitment control but are the building
blocks in the total Project Management scheme for planning activities, cost estimating
and control contracts.
The Commitment Plan should:

• provide details of planned RFAs for the commitments; total approved project budget, latest project estimate, total Authorised funds to date, the latest planned Value of Work Done (VOWD) phasing by year and the status of contingency covering the project
• match the project organisation
• be clearly delegated
• be aggregated into packages to form commitments

A CTR (Estimate Cost Time and Resources) sheet is drawn up for each activity. The number of activities should be
appropriate for the level of detail required but for practical purposes should be kept as
small as possible. The number of activities is determined by the level of breakdown of
hardware items as detailed in the CESM.

Asset and work structures

The EP Cost Management System (EP/COMS) provides a comprehensive structure for
the management of assets. The structure is POSC (Petrotechnical Open Software Corporation)and STEP compliant which is
essential for data transfer between contractors and the Company. All work breakdown
structures commitment plans and contract pricing mechanisms should conform to its
asset structures coding systems. Cost control, commitment control and expenditure
reporting are based on the activity code.This code number is identical to the finance
account number for the same commitment, which allows technical and progress
reconciliations with financial reporting. Invoices are identified with the activity code and
so charged in the financial information system.
Structure
The activity code will normally be structured from eight alphanumerical characters.

Commitment control

The following minimum requirements should be observed to be assured of sound project
financial control: Prior to commitment:

• the project must be covered by one or more approved budget items
• management must have approved the project execution plan, or, for the initial commitment plan, the development plan
• the commitment plan should define commitments in terms of hardware/function breakdowns
• the cumulative value of Authorised RFAs must always be within the total amount of the approved budget (overrun allowance may be released only by management if required)
• the commitment value must always be less than, or equal to, the corresponding RFA.