Object oriented modeling: Retrospective systems information model for constructability assessment

Automation in Construction 71 (2016) 359–371

Contents lists available at ScienceDirect

Automation in Construction journal homepage: www.elsevier.com/locate/autcon

Object oriented modeling: Retrospective systems information model for constructability assessment Peter E.D. Love a,⁎, Jingyang Zhou b, Jane Matthews c, Hanbin Lou d a

School of Civil and Mechanical Engineering, Curtin University, GPO Box U1987, Perth, Western Australia, Australia School of Civil and Mechanical Engineering, Curtin University, GPO Box U1987, Perth, Western Australia, Australia School of Built Environment, Curtin University, GPO Box U1987, Perth, Western Australia, Australia d School of Civil Engineering and Mechanics, Huazhong University of Science and Technology, Wuhan 430074, China b c

a r t i c l e

i n f o

Article history: Received 7 November 2015 Received in revised form 23 July 2016 Accepted 19 August 2016 Available online 27 August 2016 Keywords: Constructability Errors Omissions CAD E&I EPCM OOM SIM QA/QC

a b s t r a c t Object oriented modeling (OOM) has become an integral part of the design process in construction due to advances in computer software. Despite these advances there remains a tendency for Computer-Aided-Design (CAD) to be used as the medium to assist in the creation, modification, analysis and optimization of Electrical and Instrumentation (E&I) systems within heavy industrial engineering projects. In this paper, a retrospective OOM (i.e., Systems Information Model (SIM)), for the E&I systems of a utility facility, which was constructed for an Engineering Procurement and Construction (EPC) contractor for the purpose of undertaking a constructability assessment prior to the commencement of construction is presented and discussed. The CAD drawings and cable schedule produced by the EPC were provided to an E&I organization to undertake a constructability assessment; errors, omissions and information redundancy were identified and quantified. The SIM model was then used to examine a tender proposal from a construction subcontractor (CS) of the EPC; discrepancies were identified and it is suggested that differences arose due to the prevailing errors and omissions. The potential use of a SIM during construction as a quality assurance/control (QA/QC) is then examined, as it is suggested that it can be used to ensure the development of an ‘As-built’ model and provide a realistic representation of the constructed asset, which safeguards its integrity for operations and maintenance. © 2016 Elsevier B.V. All rights reserved.

Introduction “We cannot solve our problems with the same thinking we used when we created them” (Albert Einstein) Object oriented modeling (OOM) has become an integral part of the design and engineering process in construction due to advances in computer science which have resulted in an array of software tools and platforms being developed specifically for the construction industry (e.g., [1, 4,5,10,19,23,26]). In the context of construction, OOM is the digital representation of an asset as a collection of objects that contain a hierarchy of stored values for variables. This hierarchical structure means that information about each type of object needs to only be defined and stored once, as this information is ‘inherited’ by all the individual instances of this type. As a project progresses from design through to implementation the data attached to each object evolves to become more detailed,

⁎ Corresponding author. E-mail addresses: [email protected], [email protected] (P.E.D. Love), [email protected] (J. Zhou), [email protected] (J. Matthews), [email protected] (H. Lou).

http://dx.doi.org/10.1016/j.autcon.2016.08.032 0926-5805/© 2016 Elsevier B.V. All rights reserved.

and the central focus shifts toward understanding how the system will be constructed and function. Despite the advances enabled by OOM such as Building Information Modeling (BIM), within the field of electrical engineering there remains a tendency for Computer-Aided-Design (CAD) to be used as the preferred medium for design and documentation [8,9,17,18]. When CAD is used in this way there is a high risk of errors, omissions, and redundant information occurring, particularly when producing ‘As-built’ drawings [16]. In acknowledgment of the problems that may materialize by producing ‘As-built’ electrical and instrumentation (E&I) drawings using CAD, an EPC contractor approached an electrical organization to undertake a retrospective constructability assessment (i.e. to identify and eliminate errors and omissions from the design before the project commences construction). The EPC contractor was aware of the potential of errors and omissions to have materialized in the E&I drawings developed in CAD thus developed a retrospective System Information Model, which is based on OOM, to identify obstacles before their project was actually built to reduce or prevent errors, delays, and cost overruns that could potentially materialize [11]. According to Gambatese et al. [7] constructability is in part a reflection of the quality of the design documents that have been produced; if documentation is difficult to

360

P.E.D. Love et al. / Automation in Construction 71 (2016) 359–371

understand and interpret, a project will be difficult to construct. Another aim of this process was to examine how the developed SIM could be used during construction as a tool for quality assurance and control (QA/QC) to ensure that it can provide a realistic representation of the constructed asset and safeguards is integrity for operations and maintenance. It is outside the scope of this paper to provide a detailed review of the constructability literature, however, this subject has been examined extensively by O'Connor and Tucker [21], O'Connor et al. [22] Uhlik and Lores [28], Jergeas and Put [12] and Lancine et al. [3]. Systems Information Modeling A SIM is a generic term used to describe the process of modeling complex systems using object-oriented software [30]. A SIM is a digital representation of a connected system, such as electrical control, power and communication systems. When a SIM is applied to design a connected system, all physical equipment (or objects) and the associated connections to be constructed can be modeled in a database. Each object is only modeled once. Thus, a 1:1 relationship is achieved between the SIM and the real world (Fig. 1). This is in stark contrast to traditional CAD approach to producing electrical instrumentation and control systems (EICS) documentation identified in Fig. 1, which focuses on the production of drawings where an object can be represented on many drawings (m:n). Within the extant literature, there has been a paucity of research that has been able to empirically quantify the benefits of adopting an OOM approach, with those that do exist focusing on the benefits relating to the management of geometry. For example, it has been found that the adoption of OOM in the structural engineering design and detailing of reinforced concrete building structure can provide a productivity improvement of 15% to 41% [25]. As a result of this research, Sacks and Barak [25] suggested that such improvements may result in a decline in the role of drafting staff. However, structural elements are represented as geometric objects, which in an OOM approach will have embedded parametric relationships [27]. However, EICS do not possess geometric components, with the exception of cable trays, cabinets and the like, and thus the benefits of an OOM approach in this domain remain untested. As an illustration, the creation of three-dimensional

(3D) model of cabling as noted in Figure 2, for example, would be an impossible undertaking. To demonstrate the efficiency and effectiveness of using a SIM, Love et al. [16] examined the ‘As-built’ documentation that had been produced for EICS for a Stacker Conveyor system. Analysis of 106 ‘Asbuilt’electrical drawings and a cable schedule revealed a variety of documentation errors manifested themselves as labeling mistakes, inconsistent labeling, drawing omissions, missing labels, wrong design and incorrect connections. Omissions from drawings and the cable schedule accounted for 93% of all errors identified. It was revealed that a total of 803 extra man-hours would have been needed to address the omissions. In the case of all documentation errors at a total of 859 extra man-hours would be required. Love et al. [16] observed that there was considerable information redundancy contained within the 107 electrical documents. For example, 357 items appeared twice on documents with as many as 42 items appearing five times. The creation of the information redundancy contained within the 107 documents equated to an additional 598 man-hours. The Stacker Conveyor's ‘Asbuilt’cable schedule was used to create a SIM to examine how it would eliminate documentation errors and information redundancy. The average time to produce a single drawing out of SIM was two hours compared to the estimated 39 h using CAD. By using a SIM Love et al. [16] revealed that a 94% cost saving and improvement in productivity could have been attained. A SIM can be created using software such as Dynamic Asset Documentation (DAD) and applied throughout a project's lifecycle [30]. The practices associated with asset management, for example, comprise of a set of data-intensive decision-making processes, which are undertaken throughout all stages of a project's lifecycle. The development of an asset management system commences by developing a database to store and manage asset data at the beginning of a project. Yet, current practice focuses on obtaining information at the end of a project, which is expensive and time-consuming to undertake. With a SIM, data can be entered during design, construction and commissioning using the structure identified in Figure 3. Entering data into the SIM throughout each stage of development within a project enables asset owners to leverage the benefits associated with productivity and data integrity (Fig. 4).

CAD-based environment

SIM-based environment Fig. 1. The shift from CAD to a SIM-based environment.

P.E.D. Love et al. / Automation in Construction 71 (2016) 359–371

361

Case study

Fig. 2. Electrical cabling.

The retrospective creation of a SIM for asset management from existing ‘As-built’ documentation can be conducted not only for QA/ QC purposes but also to evaluate the integrity of what has been installed. The retrospective creation of ‘As-built’ documentation has been typically focused on facilities that have already been constructed using technology such as terrestrial laser scanning to obtain point clouds that can be used to develop a BIM for operations and maintenance [13,29]. As noted above, EICS do not possess geometry and therefore laser-scanning technology is inappropriate for retrospectively creating of ‘As-built’ documentation. In examining the nature of the problem with ‘As-built’ documentation for ECIS Love et al. [18] retrospectively constructed a SIM for a Copper Smelter Plant due to operations and maintenance being hindered by inaccurate information contained on the ‘As-built’ drawings. During the creation of the SIM it was discovered that unarmored rather than armored field cables had been documented and installed, which had the potential to adversely impact production targets. Moreover, it was identified that a series of electrical motors did not have their frames grounded to earth; the existence of such design errors jeopardized the integrity of the asset and there was potential for someone to be electrocuted.

A case study can be used to examine issues such as ‘why’ and ‘how’ and acquire an understanding about ‘practice’ (i.e. the actual activity, events or work) [6]. As a result, a case study can provide practical insights about industry specific problems that are being addressed and therefore enable learning and changes in practice to occur. A case study approach was therefore adopted to contribute to the paucity of empirically based research to demonstrate and justify the need to shift from a CAD to a SIM-based environment for EICS. The case study demonstrates how a SIM was retrospectively constructed and used for undertaking a constructability assessment. According to Gambatese et al. [7] constructability defines “the ease and efficiency with which a facility can be constructed. Constructability is in part a reflection of the quality of the design documents; that is, if the design documents are difficult to understand and interpret, the project will be difficult to build” (p.5). For the purposes of this research Gambatese et al. [7] description of constructability is drawn upon to provide the following operational definition: the process of examining the quality of design documents by determining the extent of errors, conflicts and information redundancy. To acquire an understanding of the case study's context and documentation that was made available to the researchers, a triangulated approach was adopted to overcome problems associated with bias and validity [24]. Cohen and Manion [2] define the process of triangulation as an “attempt to map, or explain more fully, the richness and complexity of human behavior by studying more than one standpoint” (p.254). Thus, multiple viewpoints were obtained from within the engineering organization charged with creating a retrospective SIM and constructability assessment, so as to obtain a balanced understanding of the design documents. Unstructured interviews and observations were undertaken to seek explanation and clarification about issues that were identified from the documentary sources provided, which enabled intrinsic biases that may have arisen to be overcome. Project background A Client has initiated, with the support of their Government, a US$27 billion economic development on the southern coast of the Red Sea. As part of this development an oil refinery (US$4 billion), port terminals

Fig. 3. Data structure of a SIM.

362

P.E.D. Love et al. / Automation in Construction 71 (2016) 359–371

Fig. 4. SIM and the project life cycle.

(US$1.4 billion) and a 2400 MW power plant (US$2.5 billion) are required. The refinery will cover an area of 12 km2 and have a processing capacity to process approximately 400,000 barrels per day of Heavy and Medium crudes to produce gasoline (75,000 bpd), ultra-low-sulfur diesel (100,000 to 160,000 bpd) and fuel oil (160,000 to 220,000 bpd). The planned completion date for the Refinery Plant and the Terminal is 2017. An EPCM organization was awarded a contract by the client to provide Front-End Engineering and Design (FEED), and Project Management Services (PMS). When the FEED was complete the project was divided into a total of 13 EPC packages, which included a marine terminal, utility facility and building; sour water stripper unit and amine

regeneration unit; naphtha and aromatics (benzene and paraxylene) units; two tank farms packages; crude distillation and vacuum unit; hydrocracker and diesel hydro-treater packages (Fig. 5). An electrical contractor (EC) was contacted by the EPC responsible for delivering the utility facility and building facility to undertake a constructability analysis of the existing electrical systems that had been designed and documented using CAD and retrospectively construct a SIM prior to the commencement of construction. This was required to identify errors and omissions and determine the potential of using a SIM during construction for QA/QC purposes and ensure the reliability of the ‘As-Built’ model.

Fig. 5. Structure for the project.

P.E.D. Love et al. / Automation in Construction 71 (2016) 359–371

363

Fig. 6. Application of SIM throughout a project's life-cycle.

Retrospective construction of the system information model Two methods can be used to construct a SIM using software such as DAD: 1) manually; and 2) automatically. The manual method is appropriate for new projects or where a complete cable schedule is not available. In such circumstances, engineers are required to manually create a digital model of each real-world component and cable within the SIM to form a connected system. If complete cable schedules are available, then the modelling process is considered to be straightforward using software such as DAD, as it is equipped with a function that can generate a SIM automatically based on the information derived from the cable schedules. Fig. 6 illustrates how a SIM can be applied to the throughout a construction project's life-cycle. Project information and data can be

gradually incorporated into the SIM as a project moves through its various phases to form an integrated information system that can be used for asset management. When SIM is applied to an existing project, engineers are required to construct the SIM retrospectively from data that is made available to them. Such data typically are CAD based drawings cable schedules, manuals and spreadsheets. To interpret the information and recover the data from the existing documentation, an engineering team with profound expertise is necessary. Noteworthy, there is a tendency for errors and omissions to be present in such documentation, which consequently may influence the information integrity of the SIM model. To identify and correct such errors and omissions requires a considerable amount of effort as engineers have to interpret the intent of the design and thus piece together the documentation to create a SIM.

Fig. 7. Component types and locations.

96

178 4324 14 285

23 –

– 96 – 4

– –

– 416 – 17

– –

– 500 16 157

29 –

– 126 – 67

– –

– 32 172 912

– –

1 29

– 4

– 124 23 261 – 16

4 2

– 20 2 14

6

42 148 – 64

– 11 –

23 227

17

–

21 4 470 199

199 – – – 17 8 7 16 24 – 61 – 11 – – – –

–

–

–

26

–

29

1611 – 76 – 273 – 344 – 87 26 420 68 – – – –

–

–

–

5

–

244 68

673 208 – – 7 – 2 24 10 15 47 44

95

6

8

7

58

72

20

24

26

–

–

427 35 – – – – – 32 – – 7 48 –

109

–

10

5

118

46

16

1

–

–

869 4 20 4 113 7 103 28 9 20 281 58 2 – – 20

–

–

–

12

6

108 74

1 – – – – – 1 5 1 1 14 30 2 10 4 2 5 6 37 81 7 13 10 7 9 23 – – – – 1 1 – – – – – – – – – – – –

Nitrogen plant Refinery air system Refinery steam condensate Refinery electrical system Refinery utility water Refinery cooling water Refinery fire water system Central control building Other facilities Total

Motor LCS Swbd Vender Ethernet Terminal Transformer way package Light Cabinet Panel switch MCC JB Type location

Table 1 Distribution of components from the SIM.

92 179

P.E.D. Love et al. / Automation in Construction 71 (2016) 359–371 Gas Other Transmitter Heater Gauge Switch RTD Positioner Valve detector Accelerometer devices Total

364

The engineers from the EC used the cable schedules and CAD drawings supplied by the EPC to retrospectively construct a SIM model. The cable schedules were first used to construct the frame of the model using the ‘Cable Schedule Importation’ function contained within DAD. This enabled the information contained within the cable schedules to be automatically extracted and inputted to form a digital SIM. Physical electrical devices were modeled as components and cables as connectors. Design information associated with each component and cable were also extracted and attached to the SIM. On completion of the SIM's frame, components were categorized according to their ‘Type’ and ‘Location’ attributes, i.e., functionality and physical location in the plant (Fig. 7). Connectors were classified for example according to their power load, material and size. The SIM frame is then reviewed against the information documented on the CAD drawings to verify the correctness and integrity of the model. By doing this, errors and omissions buried in the design documents can be identified and a complete SIM model can be obtained. When the modeling was complete it was found that the SIM consisted of 4324 components and 6780 connectors. The distribution of components among various types and locations are presented in Table 1. Noteworthy, as the SIM is modelled directly from the cable schedule in conjunction with the CAD drawings, to form a complete digital 1:1 realization of the original design. Acquiring data from the SIM is significantly easier than interpreting the information from paper documents. For instance, determining the quantity of materials and components for electrical systems from paper documentation is a challenging task when compared to using a SIM where all the components, connectors and their attributes are digitally modeled. Components and connectors are further categorized according to their locations and types. The use of the building ‘Search’ function within DAD enables users to verify the quantity of components, connectors, and terminals. For example, to verify the number of Junction Boxes (JB) used in the ‘Refinery Steam Condensate’ facility shown in Table 1, the ‘Type’ and ‘Location’ for selected components can be automatically calculated (Fig. 8). If the length of a particular cable type needs to be determined, then the connector search function can be used. This function will generate a spreadsheet highlighting the detailed information for the selected cables and its total length is then calculated (Figure 9). Constructability assessment The EPC provided the CAD enabled design documents to a construction subcontractor (CS) who subsequently provided a tender to undertake construction, which included a Bill of Quantities (BoQ). Noteworthy, the constructed SIM's BoQ provided to the EPC significantly differed from that identified in the construction subcontractor's tender. To identify these differences an ‘Extraction–Transformation–Loading’ (ETL) (i.e., the function performed when pulling data out of one database and placing it into another of a different type) process was undertaken and data in flat files were organized to identify them (Table 2). Moreover, it can be seen in Table 2 that the Electrical Inside Battery limit (ISBL) and Electrical Outside Battery limit (OSBL) account for over 70% of the total construction cost. Table 3 illustrates the breakdown of the costs for the Electrical ISBL and OSBL; notably work relating to cables accounted for 68% of the total cost of these packages with power cable pulling being the most significant. As a result, there is a need for cost management during construction to focus on the activities and quantities associated with the cables. While creating the SIM, a significant amount of errors and omission were identified in the CAD drawings (Table 4). The technical issues relating to cables and components accounted for 943(86.8%) and 144 (13.2%), respectively. It was observed that 152 and 268 electrical cables (38.7 km and 183 km in total) were designed with inconsistent lengths and size, respectively. A total of 99 external building cables, which totaled 48.6 km in length, were not equipped with steel wire armor.

P.E.D. Love et al. / Automation in Construction 71 (2016) 359–371

365

Fig. 8. Verification of the number of components.

Fig. 9. Verification of the length of cables.

Moreover, 336 power cables (870 km in length) did not a have protective sheath specified. Cables of this nature would invariably be exposed to extreme weather conditions (e.g., heat), corrosive materials and potential damages. Therefore it is critical for them to be equipped with

appropriate protection so as to reduce their frequency of maintenance and the plant's shutdown times. Design errors can adversely impact the procurement activities as orders may be placed for the wrong type/length of cables, which can also

Table 2 Comparison of the CS generated and SIM quantities. E& I package

Total qty (CS)

Total qty (SIM)

Construction MHs

Construction cost $USD

Percentage of construction cost

CRMS Telecom ISBL Telecom OSBL Electrical ISBL Electrical OSBL Instrumentation (COMP) Instrumentation (ISBL) Instrumentation (MPS) Instrumentation (OTHER) Instrumentation (STG) Instrumentation (STMP) Total

105 79,285 435,226 965,482 1,531,830 27,727 390,379 54,157 9,089 11,688 298 3,505,266

105 79,285 435,226 965,476 1,531,829 27,728 390,379 54,156 9,088 11,688 298 3,505,258

4,915 69,760 158,334 794,132 961,881 22,074 363,655 26,614 3,663 12,423 4,908 2,422,359

49,200 814,430 1,956,244 7,948,541 9,625,114 228,048 3,657,961 274,744 37,126 132,672 49,125 24,773,204

0.20% 3.29% 7.90% 32.09% 38.85% 0.92% 14.77% 1.11% 0.15% 0.54% 0.20% 100%

366

P.E.D. Love et al. / Automation in Construction 71 (2016) 359–371

Table 3 Summary of electrical ISBL and OSBL costs $USD. Classification

Lot

m

Panel

Pcs

Set

Total

%

Cable Cathodic protection E&I cable tray Grounding system Lighting system Substation equipment Total

– 278,478 – – – 1114 279,592

11,455,301 – 1,464,475 112,156 611,327 229,861 13,873,120

– – – – – 416,048 416,048

8815 – – 30,068 2,179,996 – 2,218,879

447,610 – – – – 338,407 786,017

11,911,726 278,478 1,464,475 142,224 2791,323 985,430 17,573,656

67.78% 1.58% 8.33% 0.81% 15.88% 5.61% 100%

influence the progress of construction activities and the project's schedule. Incorrect cable types can lead to material waste, as they would be simply redundant to the needs of the project. Without doubt this would delay the schedule, as new orders would need to be placed; productivity is impacted, costs increase, which can place unnecessary strain of project teams, as they may need to re-plan their work activities. Alternatively, if the incorrect cables are connected, then they will need to be dismantled and rework will be required. With the unit price of electrical cables varying from US$4/m to more than US$9/m depending on its type the direct cost of an error can be significant, with those of an indirect nature possibly being up to six times the rectification cost if rework [14]. Errors and omissions have also been identified on components (Table 4). For example, several motor heaters were unnecessarily equipped to small motors. A total of 90 cables had been connected between relay panels and motor control centers, however, no motor can be found to the corresponding cables. Furthermore, 10 local junction boxes had no cable connected to them.

Conflicts between the CS's and SIM generated data A significant amount of conflicting information was discovered between the data contained in the CS's tender and that generated from the creation of the SIM using the cable schedule. Table 5 presents a comparison of the results for the cable type and length. The cables have been classified into four categories: (1) Control; (2) Field Bus; (3) Instrument; and (4) Power. The power cables are further subdivided into different types according to their load classes. Total lengths of each type of cables are calculated and the average unit prices are provided based on available information. From Table 5 it can be seen that some information of the control cables and field buses was omitted from the CS's tender proposal. A comparison was also undertaken between instrument types and their quantity to verify the information consistency of the CS's data (Table 6). Here it can be seen that a total of 2846 instruments were identified in the CS's documentation. However, only 2017 instruments were identified when the SIM was created. For example, there were 75 ‘Guided wave radar level sensors’ in the tender, but there were zero identified during the creation of the SIM. Furthermore, it was observed that 202

‘Resistance Temperature Detector’ (RTD) were identified by CS whereas 458 were recognized by the SIM. Considering the above findings, rigorous QA/QC procedures need to be implemented during construction for E&I systems. Moreover to accommodate the schedule demands and cost constraints, progress monitoring during construction needs to be accurate and reliable to ensure what is actually installed is correct and documented. Quality assurance/control The Client has deployed a Quality Management Information Systems (QMIS) as its standard for inspection and QA/QC, and managing information. For example, Requests for Information (RFI's) are classified as: (1) Inspected and Accepted; (2) Job Not Ready; (3) Violation; (4) On Hold; (5) Work Not Inspected; and (6) Accepted. Surveillance logbook entries in QMIS are classified in three categories: (1) Violation; (2) Pro-active; and (3) General Comments. The QMIS logged RFI initiation records include reference to the Equipment ID for inspection, which requires manual entry. It is a mandatory requirement for the EPCM and EPCs to modify their systems to accommodate this framework. This has led to an overly complicated multi-tiered approach to QA/QC, which may lead to interoperability issues being experienced and data losses being incurred. The EPC utilizes its own application for QA/QC, which is an integrated framework that accommodates a variety of functions such as material controls and batching. The EPC has committed to incorporating the Client's QMIS into their framework. This integration required a detailed workflow to be developed to capture data as well as the integration of an array of project details such as equipment, schedules, procurement, and deliveries. However, its use requires yet another tier of integration and exchange of data for synchronization purposes. SIM QA/QC framework for construction A key advantage of using a SIM ‘Construction Module’ to perform QA/QC is that the design has now been digitally modeled, which enables a user to have quick and automatic access to detailed design information. The ‘Filter’ function provides users with the ability to search and inspect the design information using various criteria such as name, attribute, event/date, and location/type (Fig. 10). The searching results

Table 4 Technical issues identified from design. Item

Type

Number

Length (km)

Comment

1 2 3 4 5 6 7 8 9 10 11 12 13

Inconsistence Inconsistence Error Error Error Error Labeling mistake Error Error Incomplete Omission Omission Omission

152 268 99 12 76 336 16 4 5 7 90 12 10

38.7 183 48.6 1 3.8 870 N/A N/A N/A N/A N/A N/A N/A

Cable length inconsistent. Cable size inconsistent. Steel wire amour missing from external cable. Internal cable with steel wire armor. Cable connected to wrong MCC. External single core 12/20 kV and 6/10 kV cables do not have protective sheath. The Grounding Resistor tagging does not match the associate transformer. Large motor without heater. Small motor with heater. Motor heater circuit design incomplete. Motor missing. Motor with RTD is not listed on the SPI schedule. Local JB without cable.

Total

943

144

P.E.D. Love et al. / Automation in Construction 71 (2016) 359–371 Table 5 Comparison data of cable type and length. CS data Type

Length (m)

Instrument TBD Field bus Tbd Instrument Tbd 0.6/1 kV 6/10 kV Power 12/20 kV 132 kV Total length (m)

N/A N/A N/A 340,499 359,155 205,998 782,370 77,460 1,765,482

Control

SIM data Average unit price ($USD) N/A N/A N/A N/A 6.14 6.23 6.78 7.48

Total length (m) N/A N/A 340,499 1,424,983

Length (m) 32,440 172,450 96,645 11,905 358,616 174,099 832,113 84,840 1,763,108

Total length (m) 204,890 96,645 11,905 1,449,668

Table 6 Comparison of instrument type and number. EPC tender

Number

SIM

Number

Analyzers Control valve Electrical field transmitter Electrical local indicators Flow orifice Guided wave radar level sensor Magnetic flow meter Motor operated valves On – off valve Orifice plate Pressure gauge Pressure regulator Pressure relief valve RTD Temperature indicator Temperature transmitters Thermowell Total

13 136 560 182 42 75 69 319 32 158 268 8 106 202 136 202 338 2846

Analyzers Control valve Electrical field transmitter Electrical local indicators Flow transmitter Smart positioner Accelerometer Motor operated valves Motor RTD Pressure gauge Pressure control valve RTD Temperature indicator Temperature transmitter Unknown

13 102 540 71 175 17 96 284 42 53 30 458 2 184 4

Total

2071

can be displayed through a spreadsheet with all the required information for each item in the spreadsheet (Fig. 11). The spreadsheet can also be published as ‘excel’ file to be used by third party software. The ‘Review Module’ contained within DAD, for example, enables a user to review and inspect the design for each single piece of component or cable by using the ‘Pass’ or ‘Fail’ option. Comments can be attached to the items being reviewed. The review activity can be fully conducted in the digital SIM through a more effective and efficient method rather than using on paper-based drawings. Errors and

367

omissions can be identified and rectified before commence of construction thus reducing the likelihood of rework. The E&I construction scope for the entire project consists of thousands of components and millions of meters of cable/wire; the E&I scope of construction is a complex undertaking. Multiple activities have to be planned, executed and monitored for each component and cable. Traditional project management scheduling tools (e.g. critical path method) cannot be used to plan, manage, and control the thousands of activities involved. They can only present the outcome of categorizing, organizing and executing activities for E&I packages. Having an up-to-date information and undertaking accurate assessments of progress during construction requires a different and flexible method that can accommodate and respond in a timely manner. The SIM Construction Module can accommodate the complexity of E&I construction as it is: • configured to enable Plan–Do–Check–Act (PDCA) Methodology; and • generated and managed in such a way that accurate progress reports can be extracted where progress weight factors are allocated to activities (Table 7). The SIM ‘Construction Module’ is fully configurable to include QA/QC activities using simply ‘check’ type entries versus data entries. Inspection Checklists can be configured easily in a flexible way by including all the activities required by the client's QMIS framework (Fig. 12). Measuring of actual progress can be a complex process as the multi-dimensional nature of progress involves a plethora of activities that include: (1) Quantities; (2) Activities; (3) Disciplines; (4) Work Classes; (5) Steps; (6) Plant Areas or Units; and (7) Types of Equipment. The client defines measurement of construction progress in the context of a specific framework. This framework guides payments to contractors and subcontractors. Therefore, a consistent way of measuring progress needs to be established. The SIM ‘Construction Module’ can be used to determine actual progress for the E&I construction scope of work basing on work classification and weight factor allocation. Table 7 shows a typical example of classification and weight factor allocation of cable works. Example: 132 kV power cable pulling progress calculation This section demonstrates the calculation example of the first two steps of the 132 kV power cable pulling, i.e., steps 1.1 and 1.2 in Table 7. Generally, power cable pulling progress is determined as a function of total cable length and work step completions. For this example, the total length of the 132 kV power cables for pulling, derived from the SIM, is 84,840 m (Fig. 13).

Fig. 10. Information filter.

368

P.E.D. Love et al. / Automation in Construction 71 (2016) 359–371

Fig. 11. Result spreadsheet.

According to the CS's tender, the relative ‘Weight Factor for the Cable Pulling Work’ class was 4.5578%. Consequently, the weight factors of the above detailed steps are transformed in the sequence identified in Table 8. This means that at the time that activities 1.1 and 1.2 are completed, the cumulative progress will be equal to 0.228% + 2.963% = 3.191%. To assign the ‘Work Class and Step Weight Factors’ to each individual cable, the Steps and percentage of Completion for Cable Pulling are adopted from Table 8. The ‘Work Classes’ involved for Cable Pulling need to be

defined in the SIM ‘Construction Module’ as Categories shown in Fig. 14 below. The two ‘Weighted Factors for the Handling of Cable Drum’ and ‘Cable Pulling’ activities also need to be assigned for the ‘Categorized Activities’ shown in Figs. 15 and 16. These activities are assigned to a virtual ‘Work Package’ and upon completion of the ‘Do–Check’ for these activities, the data can be analysed by a simple query. The analysis results can be reviewed in the SIM and can also be exported as ‘excel files’. As the SIM ‘Construction

Table 7 Weight factor allocation example. Electrical

Discipline & work classes 1 Cable pulling 1.1 Handling of cable drum 1.2 Cable pulling 1.3 Insulation test 1.4 Markers installation 1.5 Final inspection 2 Cable termination 2.1 Termination / connections 2.2 Testing & checking 3 Cable tray 3.1 Tray support fabrication 3.2 Tray support installation 3.3 Tray installation 3.4 Grounding of trays 3.5 Installation of covers 3.6 Final inspection 3.7 Conduit support installation 3.8 Preparation and pending 3.9 Conduit installation 3.10 Identification of conduit 4 Grounding system 4.1 Handling of cable drum pulling 4.2 Grounding cable 4.3 Grounding rods & pits 4.4 Grounding rods w/o pits 4.5 Grounding bars 4.6 Overhead copper conductors 4.7 EQ. to grounding network 4.8 Installation of plates 5 Lighting system 5.1 Installation of local panels 5.2 Junction boxes installation 5.3 Conduit installation 5.4 Cable laying & termination 5.5 Fixtures installation connection 5.6 Testing & checking

Step weight factor (%) 1 5 5

2 65

3 15

Total (%) 4 10

5 5

6

7

8

9

10

11

65 15 10 5 70 70 5 5

30 30 15

25

5

5

5

5

10

15

5

5

15 25 5 5 5

5 5 10 15 5

5 5

25

10

10

10

10

20

10

25 10 10 10 10 20 10 15 15

15

15

15

35

5

15 15 15 35 5

100 5 65 15 10 5 100 70 30 100 5 15 25 5 5 10 5 10 15 5 100 5 25 10 10 10 10 20 10 100 15 15 15 15 35 5

P.E.D. Love et al. / Automation in Construction 71 (2016) 359–371

369

Fig. 12. Inspection test report.

Module’ can monitor the progress of any individual equipment or cable, it can provide an accurate and objective way for determining the progress at any point in time. Discussion Despite the impetus for using OOM and emerging software tools that enable BIM, CAD is still being used as a primary mechanism to assist in the creation, modification analysis and optimization of E&I systems within large complex construction and engineering projects. A fundamental explanation for this occurrence resides in the way projects are organized (including contractual arrangement and payments) and a simple desire to not to challenge the ‘status quo’ [16]. The sheer number of contractual layers that tend to exist in large scale industrial and

resource projects, such as in the case study presented, are unable to effectively respond, adapt and accommodate changes in design information when CAD is implemented. The hierarchical arrangement of contracts providing the legal parameters of communication sitting below the EPCM's main contract with EPCs and their contractors and then subcontractors, invariably enervates effective communication process between parties. Information is often lost in translation as the context and content dissipates, especially if an RFI is raised by engineers and is pushed up the contractual chain from one organization to another; this can be a timely process before an answer is provided to the engineers. Any changes required to the documentation that forms part of the contract to those involved in this hierarchy will need to be modified and then a copy distributed to each party. In projects such as the case study, this could result in thousands of E&I drawings being issued and

Fig. 13. Example 132 kV power cable length.

370

P.E.D. Love et al. / Automation in Construction 71 (2016) 359–371

Table 8 Weighted factors calculation. Cable pulling

100%

4.5578%

1.1 Handling of cable drum 1.2 Cable pulling 1.3 Insulation test 1.4 Markers installation 1.5 Final inspection

5% 65% 15% 10% 5%

0.228% 2.963% 0.684% 0.456% 0.228%

re-issued as changes or omissions materialize. This is not only costly process, but also adversely impinges upon productivity levels and the ability of E&I contractor to meet scheduled milestones, particularly during commissioning. The creation of an object oriented SIM provides an environment for collaboration to occur throughout a project's lifecycle as information is unified to a single point of truth whereby users can use it to manage their work simultaneously [30]. For effective inter-disciplinary collaboration within a BIM enabled environment, for example, data would need to be shared and coordinated using an industry standard format. If the end use requires accurate 3D geometry then a format such as the Industry Foundation Class (IFC) could be used for collaboration. However, if the end use does not rely on graphical content, as may be the case for operations and maintenance, then coordination of information in a non-graphical format such as Construction Operations Building Information Exchange (COBie) may be more appropriate. Yet, within an EPCM environment where there are pyramids of contracts in place and a high degree of uncertainty, especially when information does not exist, parties are subjected to ‘bounded rationality’ and therefore there is a propensity for ‘opportunistic behaviour’ [15, 20]. In this instance, collaboration is hindered, which also disables the full potential of utilizing a SIM. A caveat, however, is not to develop a contracting arrangement that supports the use of a SIM, but vice versa. Thus, a SIM can readily be used with any form of contracting arrangement as its primary focus is creating a digital representation of realworld interconnected objects that collectively form a system. The EPC in this case study, recognized the potential of issues that may have confronted them, and thus proactively initiated a constructability assessment, though retrospectively via a SIM for their E&I package; for elements with geometry this is a relatively straightforward process and thus 3D visualization and 4D simulations are regularly used. The constructability assessment was also used to examine the tender submitted by the CS; the differences between BoQs were minor for power

cabling but closer analysis revealed a significant difference between the numbers of instruments. This may have occurred due to mislabelling as well as instruments appearing unnecessarily on several documents. If the EPC had agreed to the tender submitted by the CS, then this package would have unnecessarily added to the project's costs. Another issue that comes to the fore pertains to contract conditions; whether a lump sum or cost reimbursement contract is issued. The researchers were not privy to the contractual conditions within this project, however, considering the size and complexity of the project and associated risks, it is suggested a cost reimbursement (or variant thereof) would no doubt be in place. Thus, the EPC would be reimbursed for modifying and re-issuing drawings as a result of error and omissions, even though they would have been responsible for producing them. Productivity remains an ongoing problem in large-scale complex energy, resource and heavy engineering projects. To address this issue, a mindset shift is required to move from an m:n CAD enabled digital representation, to an object oriented world of 1:1 relationships [30]. Such a change in mindset can produce new results as the benefits shown by preliminary evidence of SIM use provided in Love et al. [16] can materialize. Embedding a SIM into construction enables engineers to reduce the documentation management process workload while ensuring the ‘As-built’ model is reliable thus providing the client with confidence in the asset's integrity for operations and maintenance. Conclusion At the commencement of this paper, reference is made to Einstein who advocated a different way of thinking to solve problems. Within the field of electrical engineering, there have been limited changes to the way in which drawings have been created since the production of Thomas A. Edison's ‘System of Electric Lighting’ that was printed March 22nd 1881. The switch from manual paper based systems to CAD provided an incremental shift in productivity, but the underlying problem remains; an m:n relationship prevails. Thus, during the creation of a drawing the propensity for errors and omissions increases, as an object may need to appear on several drawings. Previous research has demonstrated that the switch from the use of CAD to an object oriented view of the world using SIM can provide transformational cost savings and productivity improvements, though these benefits have been retrospectively identified. In this paper, an EPC acknowledged the potential of using a SIM, and thus examined how it could be used as a tool for undertaking a constructability assessment of an E&I package prior to commencing construction. A SIM was generated from the CAD documentation

Fig. 14. Define work classes.

P.E.D. Love et al. / Automation in Construction 71 (2016) 359–371

371

Acknowledgments The authors are grateful to the E&I organization and its employees for providing us to such rich information and for participating in the research. The authors would also like to acknowledge the financial support provided by the Australian Research Council (DP130103018). References

Fig. 15. Weight factor for handling of cable drum.

provided, and during this process a significant amount of errors and omissions were identified; for example a total of 336 km of single core 12/20 kV and 6/10 kV cables were not specified a protective sheath. The SIM was then used to compare its quantities with those of a tender proposal from a constructor subcontractor; significant quantity differences were identified for instrumentation. Being able to assess the quality of the CAD documentation by undertaking a constructability assessment and examine the differences between the SIM and CS's tender proposal provided the EPC with confidence in the SIM's capability. Thus, this provided the impetus for examining how it can be used during construction as a mechanism for QA/QC to ensure the accurate and reliable ‘As-built’ would be developed for operations and maintenance activities. Negotiations are being undertaken at the time to determine the scope and use of the SIM during construction. Within a BIM environment, for example, its translation to construction is in its infancy and thus research examining the use of a SIM for E&I would add to this fertile area of research. The adoption of SIM requires a change in mindset from CAD to object-orientated environment so that productivity improvements can become a reality rather a vision.

Fig. 16. Weight factor for cable pulling.

[1] M. Alshawi, J. Underwood, Applying object-oriented analysis to the integration of design and construction, Autom. Constr. 5 (2) (1996) 105–121. [2] L. Cohen, L. Manion, Research Methods in Education, fifth ed. Routledge, 2000. [3] D. Lancine, G. Guoping, G. Changsheng, Adoption of building information modeling for construction project effectiveness: the review of BIM benefits (2016), Am. J. Civ. Eng. Archit. 4 (3) (2016) 74–79. [4] E.W. East, Construction Operations Building Information Exchange (COBIE): Requirements Definition and Pilot Implementation Standard. DTIC, 2007 ([Online], (Available: http://handle.dtic.mil/100.2/ADA491932. Accessed: 28th September 2015)). [5] S.L. Fan, H.P. Tserng, M.T. Wang, Development of an object oriented scheduling model for construction projects, Autom. Constr. 12 (3) (2003) 281–291. [6] B. Flyvbjerg, Case Study, Norman K. Denzin and Yvonna S. Lincoln, eds., The Sage Handbook of Qualitative Research, fourth ed.Sage, Thousand Oaks, CA 2011, pp. 301–316. [7] J. Gambatese, J. Hinze, M. Behm, Investigation of the Viability of Designing for Safety, The Center to Protect Workers' Rights, Rockville, MD, US, 2005 (Available at: http:// www.cpwr.com/sites/default/files/publications/krgambatese.pdf). [8] A.S. Hanna, F. Boodai, M. El Asmar, State of practice of building information modelling in mechanical and electrical construction industries, ASCE J. Constr. Eng. Manag. 139 (10) (2013), http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000747. [9] A. Hanna, M. Yeutter, D. Aoun, State of practice building information modelling in the electrical construction industry, ASCE J. Constr. Eng. Manag. 140 (12) (2014), http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000893. [10] H.R. Hiremath, M. Skibniewski, Object-oriented modeling of construction processes by unified modeling language, Autom. Constr. 13 (4) (2004) 447–468. [11] IPNEZ, Constructability, Practice Note 13. April, 2008, The Institution of Professional Engineers New Zealand Incorporated (IPENZ), 2008. [12] G. Jergeas, J. Put, Benefits of constructability on construction projects, ASCE J. Constr. Eng. Manag. 127 (4) (2001) 281–290. [13] J. Jung, S. Hong, S. Jeong, S. Kim, H. Cho, S. Hong, J. Heo, Productive modeling for development of as-built BIM of existing indoor structures, Autom. Constr. 42 (2014) 68–77. [14] P.E.D. Love, Auditing the indirect consequences of rework in construction: a case based approach, Manag. Audit. J. 17 (3) (2002) 138–146. [15] P.E.D. Love, P.R. Davis, S.O. Cheung, Z. Irani, Causal discovery and inference of project disputes, IEEE Trans. Eng. Manag. 58 (3) (2011) 400–410. [16] P.E.D. Love, J. Zhou, C.-P. Sing, J.T. Kim, Documentation errors in instrumentation and electrical systems: toward productivity improvement using System Information Modelling, Autom. Constr. 35 (2013) 448–459. [17] P.E.D. Love, J. Zhou, J. Matthews, C.-P. Sing, O. Olatunji, B. Carey, State of practice of building information modelling in the electrical construction industry: discussion, ASCE J. Constr. Eng. Manag. 142 (3) (2016), http://dx.doi.org/10.1061/(ASCE)CO. 1943-7862.0001069. [18] P.E.D. Love, J. Zhou, J. Matthews, C.P. Sing, Retrospective future proofing of a Cooper Mine: quantification of errors and omissions in ‘As-Built’ documentation, J. Loss Prev. Process Ind. 43 (2016) 414–423. [19] S. Malsane, J. Matthews, S. Lockley, P.E.D. Love, D. Greenwood, Development of an object model for automated compliance checking, Autom. Constr. 49 (2015) 51–58. [20] P. Mitropoulos, G. Howell, Model for understanding preventing and resolving project disputes, ASCE J. Constr. Eng. Manag. 127 (3) (2001) 223–231 (2001). [21] J. O'Connor, R. Tucker, Industrial project constructability improvement.", ASCE J. Constr. Eng. Manag. 112 (1) (1986) 69–82. [22] J. O'Connor, S. Rusch, M. Schulz, Constructability concepts for engineering and procurement, ASCE J. Constr. Eng. Manag. 113 (2) (1987) 235–248. [23] F. Ozel, Object oriented representation of design decisions, Autom. Constr. 8 (12) (1998) 24–37. [24] M.Q. Patton, Qualitative Evaluation and Research Methods, 2 ed. C.A: Sage, Newbury Park, 1990. [25] R. Sacks, R. Barak, Impact of three-dimensional parametric modeling of buildings on productivity in structural engineering practice, Autom. Constr. 17 (4) (2008) 439–449. [26] Z. Turk, Object-oriented modelling and integrated CAD, Autom. Constr. 1 (4) (1993) 323–337. [27] J. Underwood, U. Isikdag, Handbook of Research on Building Information Modelling and Construction Informatics: Concepts and Technologies, IGI Global, 2010. [28] F. Uhlik, G. Lores, Assessment of constructability practices among general contractors, ASCE J. Archit. Eng. 4 (3) (1998) 113–123. [29] R. Volker, J. Stengel, F. Schultmann, Building information modeling for existing buildings—literature review and future needs, Autom. Constr. 38 (2014) 109–147. [30] J. Zhou, P.E.D. Love, J. Matthews, B. Carey, C.-P. Sing, An object oriented model for life cycle management of electrical instrumentation control projects, Autom. Constr. 49 (2015) 142–151.