Estimating for the Process Plant Industry - Leveraging Today's Computer Technology Authors: Trevor Staples, Manager Estimating, Foster Wheeler Energy Limited Tim Mote, Global IT Project Manager, Foster Wheeler Energy Limited Dr. Peter Sanio, CTO, RIB Software AG

Abstract This paper examines the roll-out of a major new estimating system within Foster Wheeler. This project has taken three years from initial conception to successful implementation, demonstrating not only the considerable technological step-forward achieved but also the significance of the change involved. Firstly, this paper examines why the process plant contracting industry needs to enhance its IT platforms. Specifically it will consider the need to produce estimates to the required accuracy while at the same time reducing the time and cost associated with estimate production, improving estimating integrity through better quality control, and enhancing estimate transfer into project execution. The second part of the paper explains and demonstrates how these improvements have been implemented in Foster Wheeler. Key areas to be explained include how the system was configured on a worldwide operational basis and the implementation of advanced estimate modelling techniques. Finally, the paper will identify the key issues related to personnel change management and training of the end users that must be considered when successfully rolling out a major new IT system. INTRODUCTION Foster Wheeler embarked on a programme to enhance its estimating processes and systems approximately three years ago. A key element was the decision to invest in an estimating database. As an international engineering and construction contracting company, it produces estimates from major centres in Europe, USA, Asia-Pacific and Africa. These estimates are critical as they not only underpin and support clients’ investment decisions, but also Foster Wheeler’s own business, especially for lump sum turnkey engineering, procurement and construction (EPC) projects.

The estimates produced cover the full range of process plants: upstream oil and gas, LNG and GTL, refining, chemicals and petrochemicals, pharmaceutical and power facilities. The projects range in size from under $1 million to multi-billion dollar complexes. Against this background, this article discusses the following four issues:

• Why invest in a new system? • FW Evaluation of in-house versus third

party solutions • System implementation and capability • Key management issues

The system has been implemented by the customisation of an available database from RIB software headquartered in Stuttgart, Germany. This was a collaborative effort across the major Foster Wheeler centres. The most significant benefits were expected for estimates with accuracies between +/-10% to +/-10%, but conceptual estimates at +/-30% would also benefit. I. WHY INVEST IN ESTIMATING SYSTEMS? The first step in the process was to identify very clearly the desired objectives, prior to undertaking research into the options available to enhance the estimating system. Without a clear view on what the whole project needed to achieve, it would not be possible to credibly assess the different options available, or to judge progress at various stages of the implementation process. Creation of Knowledge Bank The first factor that led to the conclusion that investment was needed arises from the changing nature of employment. Most people have heard of the term the ‘psychological contract’ between employer and employee. This has changed significantly over the past few decades. Whereas 20 to 30 years ago employees could reasonably expect long-term employment in return for loyalty to their employer, it is now clear that no organisation can offer a "job for life". Both employee and employer must continuously adapt to change and address new business challenges. This has far-reaching implications for the estimating function.

The engineering and construction contractor must recognise that it has less easy to undertake long-term development of a group of estimating engineers such that they all develop an intuitive ‘feel’ about their estimates. As a direct result of this, active knowledge management has become more critical, as it is a risk for contractors to rely totally upon key individuals. This was recognised within Foster Wheeler and an obvious solution to consider was development of a database system to enable data capture from past estimates and completed projects. The objective was therefore to easily capture key data in a database system, which would augment the skills of the estimating engineer, who could use this data to assess the integrity of an estimate.

Figure 1 – The Estimate Life-Cycle This is shown in Figure 1, which depicts the life-cycle of the estimate. It is vitally important to build strong feedback loops between past estimates and as-built project data, into new estimates. System Integration The second critical factor was to achieve integration of different systems. It is feasible to link some systems together. However, the robustness of the links and hence the integrity of the estimate may not always meet today's quality requirements. Hence the second objective was to ensure that any new system was fully integrated with maximum electronic transfer of data. This decreases the chances of human error and offers potential productivity gains. Operating Efficiency The third critical factor was identified against the background of an increasingly competitive business environment. Automation of some aspects of the estimating process helps reduce direct job-hour costs associated with estimate production. Secondly, the ability to work-share

between different centres offered efficiency savings, facilitating the use of low cost centres and by allowing easier resource levelling. Hence two objectives were identified:

1. automate estimate production where feasible while maintaining estimate accuracy and integrity, and

2. allow work to be undertaken in different centres.

II. OPTIONS EVALUATED Prior to selecting the chosen software supplier a number of options were considered. The possibility of enhancing Foster Wheeler's current systems was discarded fairly easily. It was the intent to focus Foster Wheeler's internal efforts on the pure estimating aspects of the project rather than the computer programming issues. There then appeared to be two options for a new system. The simpler option was to buy an already-developed system that offered the capability to produce a single estimate in a database environment, hence providing the benefit of system integration. However, while this would have offered an improvement on the current systems in place, it would not meet the objectives of providing a knowledge bank for the future and allowing concurrent work from different centres. Hence Foster Wheeler chose an option that would produce a single estimate in a database environment and provide a knowledge bank: the estimating modules within RIB’s Construction Suite. This was based on an enquiry to RIB that set out the system requirements. However, as a result of choosing to advance the estimating system considerably to meet the above objectives, it was essential that RIB and Foster Wheeler collaborated during implementation to meet both parties' objectives. Prior to Foster Wheeler's involvement with RIB, RIB had not previously worked with a major engineering and construction contractor outside of its home market. Equally, while Foster Wheeler understood its key objectives it was also necessary to learn and understand how the software could be exploited to its maximum capability. Hence teamwork between both companies was essential such that RIB gained from understanding the complexity of process plant estimates generated for many world-wide locations and Foster Wheeler gained from understanding the potential of the system to meet its future business needs.

Through this teamwork some enhancements were made by RIB to its base software. These enhancements were critical to Foster Wheeler and therefore also essential for RIB if it was to be capable of marketing its product to other companies working in the international engineering and construction market. Specific enhancements included enabling true multi-currency capability and development of a feature to import/export data from other systems. III. SYSTEM IMPLEMENTATION AND CAPABILITIES SYSTEM IMPLEMENTATION The system implementation had to address three major areas:

• System roll-out • Interfaces with other systems • System Integrity

System roll-out Since the new system was to have full database functionality it was essential that all individual items within it could be uniquely identified. Like all engineering and construction contractors and their clients, Foster Wheeler has its own code of accounts and commodity codes. It was therefore necessary to import the code of accounts and commodity codes into the system. This was a major activity to establish the basic structure suitable for use within Foster Wheeler. However, once completed it enables the functionality of the database to be fully utilised. This is because the completed estimate can ultimately be analysed against any set of cost codes and detailed output generated down to commodity code level. Once the system was configured to work with Foster Wheeler’s codes it was then necessary to establish within it catalogue data that contained estimating norms associated with the codes. These norms include direct labour job-hours for installation of specific commodities and standard pricing data obtained from procurement systems. These catalogues can be updat ed and form the basis for each new estimate. Interfaces with other systems Foster Wheeler has developed over a number of years a sophisticated and proven program for piping estimates, called PIPEST. To keep the functionality of PIPEST, it was necessary to keep this as a standalone program outside the RIB database. However, to ensure that PIPEST

‘talked’ electronically to the database, it was necessary to upgrade PIPEST and move its operating platform to Oracle. This migration project proceeded alongside the database development. To maximize the extent of estimating calculation that is undertaken within the database and to minimize external links, PIPEST was reconfigured as part of this upgrade to work as a quantification tool. Hence all pricing of the quantification occurs within the database. Additionally Foster Wheeler had a program for automatic calculation of electrical cable sizes. This was therefore also enhanced to ensure there was an electronic link between the external program and the database. System Integrity On completion of the development of the database it was necessary to prove the system integrity. Currently the database is linked with some external systems as shown in Figure 2. A programme of tests was undertaken to prove the dat a transfer and validate the individual formulae within the system. This involved considerable effort, as it was necessary to carry out hand calculation checks. As an additional safeguard, the new database was run in parallel with existing systems as a cross check on a number of estimates. Once these checks were completed, it was possible to be confident that the automation was working well. .

Figure 2 - System Integration SYSTEM CAPABILITIES There were three main methods of enhancing the database, which would improve operating efficiency. These consisted of two areas of automation: templates and assemblies. The third method was to enable work-share. In addition,

data analysis with key figures provides excellent benchmarks for future use. Templates Templates are very similar to applying a set of formulas within a spreadsheet. An example is shown in Figure 3. This shows a template constructed for a typical foundation type. The estimating engineer enters key design data and then the pre-programmed template calculates the required quantities. The templates’ formulae are based on the relevant commodity codes. This means that it is possible to interrogate the estimate to a significant level of detail and use the initial estimating data to establish early project material take-off information so maximising the value gained from the estimating effort. The advantage of templates is that many standard templates can be created which enable automation of the estimating process. For example, templates have been created for the following major areas:

• Civil foundations • Equipment painting and insulation • Instrument hook-ups

Figure 3 – Example Civil Foundation Template Assemblies Assemblies offer a greater level of automation and sophistication. The example in Figure 4 shows an assembly for a wellhead. This assembly identifies that a wellhead consists of many distinct elements. Within each individual element that makes up the overall wellhead there is a series of templates. Hence in the example the wellhead consists of a burn pit that can be quantified from some key design data that is established by the estimating engineer. For each of the individual wellheads calculations are carried out automatically by the database based on the key data.

Figure 4 – Example Wellhead Assembly Work-share The database has been implemented in a number of operating centres. From each unit, it is possible to access all other centres. Additionally the database can be accessed via satellite centres over standard IT links. This is depicted in Figure 5.

Figure 5 – Configuration of Work -share Capability This capability means that a single estimate can be worked on from a number of centres. The individual responsible for the estimate can always see in real-time its current status. Clients can be allowed remote access and when presenting an estimate in a remote location access can be gained from a laptop over a secure internet link. This gives considerable advantages in being able to move work to a person’s location, rather than move the person to the work location. Data Analysis A critical objective identified in the early stages was that the database had to become a knowledge bank for the future. A key feature of the database was therefore the ability to analyse an estimate and generate key statistics. Through the report writing features within the system this capability has been established. Some of the data generated is listed below in Table 1.

Major equipment Number Concrete Total volume Volume per item Above ground piping Total linear length

Total weight Length per item Weight per item

Steelwork Total weight Weight per item Field instruments Number Number per item Instrument valves Number Number per item Cable (both instrument and electrical) Total linear length Length per item Direct labour job-hours Total hours

Total value Hours per trade Value per trade

Lang factors Ratios material and construction to equipment, total investment cost (TIC) to equipment

Engineering services Total job-hours Total cost

Job-hours per item Percentage of TIC

Construction management Total job-hours Total cost

Ratio of construction management to direct labour Percentage of TIC

Table 1: Key Estimating Statistics Generated from the Database

When an estimate is being checked by the estimating engineer or reviewed at a management level it is very easy to establish whether key relationships are making sense. This frees the skilled estimators to apply their knowledge in making the subjective or intuitive judgements that are critical to good estimates. Estimating effort is focused on areas where there appear to be discrepancies against what is expected from previous estimates and completed projects. The RIB system can search across different estimates allowing re-use of data for current estimates. For example, it is possible to search across several estimates for pricing data relating to an equipment type. This enables an estimate of an item to be generated from analysis of in-house data without necessarily having to obtain additional quotations from the market. Finally, the system allows isolation of a specific part of the estimate and re-use of individual parts of the work breakdown structure. This ability to analyse specific parts of the estimate, both through the work breakdown structure and against specific cost and commodity codes, is valuable when investigating value engineering and "what-if" scenarios. It also becomes easier to isolate the high value areas of the estimate and focus attention on these when optimising cost of major projects, for example, when applying value improvement practices. This capability is only available in an integrated database system with search capability across the cost code structure.

These features are particularly relevant when carrying out conceptual estimates. It may not be feasible due to cost and schedule constraints to obtain market quotations. However, by breaking out specific aspects of other estimates, and using the factors and statistics generated from the detailed estimates it is possible to develop use the data in applying parametric models for conceptual estimates. IV. KEY MANAGEMENT ISSUES It is apparent from the popular press that many IT projects appear to fail - high profile problems have occurred in many areas. A key lesson from this project is that the non-IT issues must be managed carefully. It is all too easy to focus on the hard issues of the IT investment, while ignoring or not allowing enough focus on the end users and their needs. This latter aspect is absolutely critical. Three key areas are worth sharing from this project. Firstly, the same critical success factors which apply to an EPC contract for example, an upstream oil and gas or pharmaceutical facility, apply here too: very clear front-end scope definition, good ongoing project management and control, and most importantly very tight management of scope change. Without the latter there is a risk that the project will extend considerably in terms of schedule and cost. And then there are the ‘human’ factors… Training on the system is essential but the key issue here is timing of that training. Some training was undertaken too early, which meant users could not move quickly from training to system

use and therefore lost the benefit of the training. This was as a result of trying to move the implementation ahead as quickly as possible. By learning this lesson early on, it was possible to train individuals ‘just in time’ i.e. just before they were about to use the system, so that they could get maximum benefit from the training. The third key area relates to the process change that is almost inevitable with implementation of a new IT system. As stated above, one of the objectives of the system implementation was to gain some productivity enhancements and to enable workshare. Consequently, major changes in IT systems should, and do, lead to changes in work processes. The incorporation of the templates and assemblies described earlier resulted in changes in the processes to be followed by the estimating engineer. However, to gain acceptance of these changes it was necessary to encourage individuals to see that they could benefit from these changes. This was achieved by getting the individual estimating engineer to be actively involved in identifying the improvements that could be achieved, to get ‘buy in’ to the changes. Where changes were implemented without this involvement, the lack of ownership by individuals was far more apparent. It is never going to be possible to involve everyone who will be affected by an implementation of this nature. Therefore by recognising where ownership issues exist, attention can be given to explaining the change and gaining acceptance - if the ownership issue is not recognised then acceptance will be much harder to achieve and ultimate success not necessarily guaranteed. V. CONCLUSIONS AND FUTURE DEVELOPMENTS Implementation is now complete and the new system is in use. The key objectives originally identified have been achieved as summarised in Figure 6.

Figure 6 – Achieving the Objectives The intent now is to populate the database with historical estimates and past project data where to deliver future value. Additionally it is necessary to continue to identify areas of estimate production where templates/assemblies can be used to gain further automation. Recognising that change and improvement are essential in today’s competitive business environment, Foster Wheeler is continuing to work with RIB to identify how the software can be further exploited. Critically under implementation now is the capability to use the system to manage construction sub-contracts from estimate to close out. Following this implementation Foster Wheeler is examining how to utilise the software during project execution to fully close the loop from estimate generation to project closeout, so further enhancing its knowledge bank in a fully integrated manner.