A Preliminary Study on the Comparative Schedule Analysis of Traditional and Advanced Work Packaging Plans for Nuclear Power Plant Construction

Abstract

The construction of nuclear power plants (NPPs) involves complex and long-duration projects where schedule delays critically affect project performance. To overcome this challenge, Advanced Work Packaging (AWP) has emerged as a promising alternative approach. It offers a more integrated and structured way to plan and execute projects, aiming to improve efficiency and reduce the risk of delays. To evaluate the potential benefits, this preliminary study developed and compared a traditional phase-based schedule and two AWP-based schedules. Delay simulations and productivity adjustments were conducted to analyze schedule resilience and mitigation performance. The results show that AWP-based schedules enhance traceability, expand work package granularity, and improve recovery against engineering delays through structured segmentation and Workface Planning (WFP). These findings quantitatively demonstrate the potential of AWP to improve scheduling efficiency not only in NPP projects but also mega construction projects while also identifying gaps in maturity, boundary definition, and integration practices that must be addressed for broader adoption.

1. Introduction

The construction of nuclear power plants (NPPs) entails large-scale, complex projects with long durations and high regulatory and safety requirements. Schedule delays in such projects often lead to significant cost overruns and stakeholder risks [1]. In the nuclear domain, these delays are especially consequential because of safety and regulatory interfaces and often cascade into severe cost overruns and extended completion times [2]. Empirical rankings identify policy changes and approval delays as dominant schedule risks, reinforcing the need for policy frameworks that prioritize design maturity and a stable, predictable safety-regulatory regime with explicit construction-readiness requirements [3,4]. Accordingly, an effective management system ensuring documentation, configuration control, and lifecycle traceability becomes a prerequisite for disciplined schedule control [5].

Advanced Work Packaging (AWP) has emerged as a structured project execution methodology that improves schedule predictability, cost efficiency, and quality through well-defined work packaging and the alignment of engineering, procurement, and construction (EPC) processes [1,6,7,8,9,10,11,12,13]. International adoption of AWP is expanding across industrial sectors (e.g., oil & gas, petrochemical, infrastructure) [9,10,11,12] and owners have begun to require bidders to incorporate AWP principles into project planning [9].

Most prior AWP research has focused on explaining concepts, pros and cons, it has rarely provided explicit schedule logic or practical comparisons with traditional practices. This study aims to clearly demonstrate the differences between AWP and traditional project management methods, specifically within the context of a nuclear power plant construction project. The research focuses on two main objectives. First, it identifies the specific changes required in the WBS, deliverables, and activities when transforming traditional schedules into an AWP-based approach. Second, it compares this AWP-based method to current Korean traditional NPP practices to pinpoint the practical advantages and disadvantages of its adoption. Ultimately, this study offers a clear, scheduling-level comparison to illustrate how AWP can practically change project planning and control, and frames this comparison in terms of schedule resilience—the ability of plans to absorb engineering delays and recover execution feasibility [14,15]. The novelty of this work lies in being the first preliminary attempt to systematically apply AWP concepts to nuclear power plant scheduling. By developing both traditional and AWP-based schedules on a common Nuclear Island model, the study provides an intuitive apples-to-apples comparison, contributing methodological insights and practical evidence to guide future adoption of AWP in the nuclear domain.

2. Background: AWP Application for NPP Construction

2.1. Previous AWP Application Research

AWP is a structured execution framework that aligns EPC through clearly defined package types and workface planning. Its core components are well established in the literature: Construction Work Areas (CWAs), Path of Construction (POC), Engineering/Procurement/Construction Work Packages (EWP/PWP/CWP), Installation Work Packages (IWP), and Workface Planning (WFP) [13]. In practice, early definition of the POC and constraint-free release of IWPs are repeatedly highlighted as mechanisms that improve schedule predictability and field productivity [13,16,17].

Across industrial sectors such as oil and gas, petrochemical, infrastructure, empirical and case-based studies report that AWP adoption is associated with productivity gains, reductions in total installed cost, and improved schedule reliability and safety performance, with benefits traced to disciplined work packaging, earlier engineering readiness, and clearer inter-package interfaces [10,11,12,13,16,17]. Methodological work has also proposed decision frameworks to evaluate AWP’s costs and benefits by structuring both direct costs and indirect/intangible benefits for investment decisions [18].

Scope extensions have been studied to connect AWP with commissioning and startup. For example, package schemas and handover criteria have been proposed to bridge IWP-level field completion to system-based commissioning activities, thereby reducing late rework and improving startup readiness [19]. On the information side, researchers emphasize bridging the engineering and construction 3D so that package boundaries, sequencing, and constraints are consistently represented across design and construction environments; poor alignment has been linked to rework and schedule slippage [20].

Implementation studies and reviews consistently surface enablers and barriers: organizational maturity and training, data/model standards, system interoperability, governance and roles/responsibilities, and stakeholder alignment. Sectoral surveys and literature syntheses report that ambiguity in implementation plans, capability gaps, and fragmented systems are recurring inhibitors that must be addressed for AWP to deliver the intended outcomes [21,22,23,24,25].

In summary, prior research indicates that (i) package-based execution with proactive constraint removal tends to improve schedule and cost performance, (ii) early POC definition and model/data alignment are necessary preconditions, and (iii) organizational readiness largely determines realized benefits [10,11,12,13,16,18,19,20,21,22,23]. However, there weren’t much research showing the detailed schedule logic and work package information.

2.2. Considerations for AWP Application in NPP Construction

NPP projects face tighter regulatory, quality, and documentation requirements than most industrial projects; they also feature system-centric testing and turnover, long-lead procurement, and complex interfaces. Translating AWP into this environment calls for the following considerations, which draw on both international megaproject insights and Korean NPP practice.

(1)

Alignment with licensing hold points and regulatory milestones.

NPP schedules are highly sensitive to front-end definition and external approvals. POC and CWP boundaries should be established to coincide with licensing hold points and inspection gates so that release of EWPs, CWPs and IWPs is feasible within the regulatory timeline [1]. In practice, policy frameworks should prioritize design maturity and regulatory stability with explicit construction-readiness requirements and a stable, predictable safety-regulatory regime, directly addressing approval-driven schedule risk [3]. Such approval delays are repeatedly ranked among top schedule drivers in empirical assessments [4].

(2)

Mapping to QA/QC and document/procedure control.

Korean NPP delivery relies on mature process and procedure management (standardized workflows, controlled documentation, rigorous revision/approval cycles). AWP packages must therefore be mapped to existing QA records and procedure hierarchies so that package completion criteria and evidence (drawings, specifications, inspection/test records) are auditable within the current quality system [26]. An effective management system—ensuring documentation, configuration control, and lifecycle traceability—is essential and AWP packaging should be embedded within the project’s safety/quality/regulatory interfaces [5].

(3)

Bridging phase-driven WBS and CWA-driven AWP.

Domestic practice commonly uses phase (EPC) and discipline-driven WBS with CPM and EVMS (WBS–Control Account–cost/schedule integration). Because AWP organizes scope by CWA, a direct 1:1 mapping gap exists. Clear WBS to CWA/package mapping rules (e.g., WBS code extensions, cross-walk tables) and redefined progress/cost roll-ups are needed to preserve EVMS integrity while introducing EWPs/PWPs/CWPs [27]. Recent Korean reviews of AWP for NPPs reach similar conclusions and recommend early decisions on package boundaries and coding to avoid late rework in controls and reporting [24,28].

(4)

Engineering–construction data/model consistency.

EWP deliverables (drawings, MTOs, specifications) and CWP/IWP readiness attributes (materials, access, safety, quality checks) should live in compatible data models to prevent interface gaps. Research urges a structured bridge between engineering and construction models so that package metadata, sequencing, and constraints are consistent and machine-readable across tools [20]. Consistent information and records reduce rework from errors and late design changes, which are frequently observed delay factors in NPP projects [2,28].

(5)

Commissioning interface (system-based turnover).

Because NPP turnover is system centric, commissioning packages and handover criteria (e.g., completeness, QA records, punch status) should be explicitly linked to CWP/IWP completion states. The AWP literature proposes package schemas and governance to tie construction work faces to system-level commissioning workflows [19,28,29].

To apply AWP to nuclear projects, management plans must span the full delivery lifecycle, from early definition to final commissioning, and should incorporate the considerations outlined above. Using the PMBOK process framework as a guide [30], this study concentrates on early-phase foundations because they set constraints for all later work and can be evaluated with available project data. Accordingly, the research scope focuses on defining the WBS and developing the schedule that is traceably linked to that WBS, which together act as prerequisite enablers for integrating engineering, procurement, and construction in an AWP context [24,27].

3. Methods: Research Framework and Model Set Up

This study implements a comparative schedule analysis comprising model configuration, development of a traditional phase-based and two AWP-based (AWP-M1, AWP-M2) schedule, parameter adoption (productivity uplift and workface-planning), and delay–recovery simulations. Figure 1 details the procedure for the comparative schedule analysis. Productivity is applied as literature-based assumptions, while explicit resource and spatial constraints are not modeled in this comparison.

3.1. Model Configuration

A scaled-down hypothetical model of the Nuclear Island (NI), comprising the Reactor Containment Building (RCB) and the Auxiliary Building (AB), was configured as the project scope. This was selected because it captures the dominant critical path of the NPP schedule and is tightly coupled with startup milestones and Final Safety Analysis Report (FSAR) interfaces; accordingly, much prior work focuses on compressing this path. In addition, construction practices within the NI & AB already employ segmentation concepts equivalent to CWAs and the POC, which enables consistent area-based breakdown and sequencing in this study. CWAs were defined for each major zone, and the POC was established according to logical sequencing (Figure 2).

The geometric reference followed Shin-Kori Units 3 and 4 Final Safety Analysis Report (FSAR), Chapter 1: General Plant Information [31]. To isolate schedule-structure effects, the model was intentionally simplified. Only structural boundaries and construction zones were represented, while detailed equipment layouts, piping systems, and functional rooms such as the Main Control Room (MCR) were omitted. For the AB, functional elements that are typically concentrated in specific zones in practice were assumed to be uniformly distributed across CWAs to minimize modeling bias and support analysis of AWP-based WBS and schedule structures. Accordingly, to assume a model that permits the uniform application of simplified, standardized durations to unit activities in the schedules developed in this study, dimensional scaling was applied to derive realistic activity durations; key dimensions and scaling ratios are summarized in

Table 1. Dimensional assumptions for the hypothetical model.

Component	Parameter	APR1400 Reference *	AWP Model *	Scaling Ratio (%)
RCB	Radius (r)	0.00 m	0.00 m	65.22%
	Area (A)	0.00 m2	0.00 m2	43.06%
	Height (S.L.)	0.00 m	0.00 m	49.18%
	Height (Dome)	0.00 m	0.00 m	54.22%
	Dome Volume	0.00 m3	0.00 m3	28.26%
	Total Volume (V)	0.00 m3	0.00 m3	22.71%
AB	Width	0.00 m	0.00 m	67.31%
	Depth	0.00 m	0.00 m	82.35%
	Area (A)	0.00 m2	0.00 m2	55.32%
	Height	0.00 m	0.00 m	43.48%
	Total Volume (V)	0.00 m3	0.00 m3	25.13%

* All dimensional values are replaced with “0” for confidentiality.

.

3.2. Traditional Schedule Development

3.2.1. WBS Creation

Figure A1 presents the traditional WBS developed based on typical Korean NPP practices [24]. The structure supported the study objectives by enabling comparison with AWP-based WBS models and by providing a foundation for schedule development. It followed a discipline-driven, phase-based hierarchy segmented into EPC phases and was optimized for progress measurement and cost management, consistent with Korean NPP practice. Unlike AWP-based WBS structures, it was not explicitly aligned with CWAs or the POC and primarily functioned as a performance-tracking structure rather than a field-execution guide.

Work packages in this study were defined at Level 4 of the WBS:

• Engineering Functional Breakdown Structure (FBS) items representing discipline-specific drawings, specifications, and calculations.
• Procurement packages (PO, Purchase order) representing major material and equipment supply.
• Construction package (CP) activity types representing grouped site work activities. The selected Level 4 items used for modeling are listed in Appendix A.4 Scheduling Assumptions.
• Table A3. Selected engineering FBS for the model.

  Discipline	FBS No. *	Description
  All/Common	4XX	System design criteria
  C	1XX	Structural drawing
  	1XX	Liner plate drawing
  	1XX	Structure detail drawing
  	2XX	Specification-reinforcing steel bar
  	2XX	Specification-containment liner plates
  	2XX	Specification-ready mixed concrete
  	3XX	Structural analysis
  	3XX	Liner plate calculation
  E	1XX	Single line diagram
  	1XX	Cable tray drawing
  	1XX	Conduit plan drawing
  	2XX	Specification-cables
  	2XX	Specification-cable trays and fittings
  J	1XX	C&ID
  	1XX	Instrument location drawing
  	2XX	Specification-instruments
  M	1XX	P&ID
  	2XX	Specification-pumps
  	2XX	Specification-crane
  	NSSS	Specification-nsss equipment
  	1XX	P&ID
  P	1XX	Piping support drawing
  	1XX	General arrangement drawing
  	1XX	Piping iso drawing
  	2XX	Specification-piping
  	2XX	Specification-valves
  	3XX	Piping stress analysis

  * All FBS numbers are replaced with “X” for confidentiality.

  ,

  Table A4. Selected PO for the model.

  PO *	Description
  CXXX	Reinforcing steel bar
  CXXX	Ready mixed concrete
  CXXX	Containment liner plates
  MXXX	Pumps
  MXXX	Hoists & cranes
  NSSS	NSSS equipment
  PXXX	Valves
  PXXX	Piping
  JXXX	Instruments
  EXXX	Cable trays and fittings
  EXXX	Cables

  * All PO numbers are replaced with “X” for confidentiality.

  and

  Table A5. Selected CP and activity type for the model.

  CP *	Description	Activity Type *	Description
  CP-AX	Major building and Related structure	AX	Basemat & slab
  		AX	Wall & column
  		AX	Steel liner plate
  CP-PX	Piping work	PX	Pipe & support
  CP-EX	Cable tray	EX	Cable tray & support
  CP-EX	Cabling and termination	EX	Cabling & termination
  CP-JX	Instrument and Control installation	JX	Instrument and Control installation

  * Part of the numbers are replaced with “X” for confidentiality.

  .

These selections enabled clear mapping across engineering, procurement, and construction sequences, supported realistic activity-duration estimation, and established a structured basis for comparative analysis between the traditional baseline and AWP-based schedules. Details on how these deliverables informed schedule logic and activity durations are presented in the following subsection.

3.2.2. Schedule Development

Define Activities

Based on the selected Level-4 work packages, activities and associated deliverables were defined together with their relationships to maintain a consistent structure across EPC phases and to identify standard sequences per WBS item. Engineering activities modeled drawings and documents as Preliminary Issue (PI) followed by Construction Issue (CI), and specifications were issued for Invitation to Tender (ITT). Procurement sequences comprised ITT; bidder selection/purchase-order award/vendor data submission (ITT/PO/VD); review and approval (R&A); and fabrication and site delivery (Fab). Construction execution was organized by physical segmentation (area, level, and type of work). Representative relationships are illustrated in Figure A7 and Figure A8.

Sequence Activities

To ensure comparability between the traditional and AWP-based approaches, detailed activity relationships were simplified and standardized. Initial sequences derived from Korean NPP cases (Figure A7) were refined into a standard sequence model (Figure A8), aligning engineering, procurement, and construction phases across disciplines and providing a practical baseline for comparative analysis. A 30-day lead time was applied between completion of engineering/procurement activities and the start of the corresponding construction tasks to reflect preparation and readiness requirements. A relationship-mapping table (* Part of the numbers are replaced with “X” for confidentiality.

Table A6. Engineering (FBS)-to-Procurement (PO) relationships.

BLDG		Sys.	Predecessor			Successor *
RCB	AB		OBS	FBS *	DESCRIPTION	CXXX	CXXX	CXXX	MXXX	MXXX	NSSS	PXXX	PXXX	JXXX	EXXX	EXXX
O	O		P	XXX	GENERAL ARRANGEMENT DRAWING	O		O	O	O	O	O	O	O	O	O
O	O	O	All	XXX	SYSTEM DESIGN CRITERIA
X	O	O	M	XXX	P&ID				O			O	O	O
O	X	O	N	XXX	P&ID						O		O
O	O		P	XXX	PIPING STRESS ANALYSIS
O	O		P	XXX	PIPING ISO DRAWING				O		O	O	O	O
O	O		P	XXX	PIPING SUPPORT DRAWING							O	O
O	O		C	XXX	STRUCTURAL ANALYSIS
O	O		C	XXX	STRUCTURAL DRAWING	O			O	O	O
O	O		C	XXX	STRUCTURE DETAIL DRAWING	O			O	O	O
O	X		C	XXX	LINER PLATE CALCULATION
O	X		C	XXX	LINER PLATE DRAWING			O
O	O	O	E	XXX	SINGLE LINE DIAGRAM										O	O
O	O		E	XXX	CABLE TRAY DRAWING										O	O
O	O		E	XXX	CONDUIT PLAN DRAWING										O	O
O	O	O	J	XXX	C&ID									O
O	O		J	XXX	INSTRUMENT LOCATION DRAWING									O
O	O		C	XXX	SPECIFICATION-REINFORCING STEEL BAR	O
O	O		C	XXX	SPECIFICATION-READY MIXED CONCRETE		O
O	X		C	XXX	SPECIFICATION-CONTAINMENT LINER PLATES			O
X	O		M	XXX	SPECIFICATION-PUMPS				O
O	X		M	XXX	SPECIFICATION-CRANE					O
O	X		M	XXX	SPECIFICATION-NSSS EQUIPMENT						O
X	O		P	XXX	SPECIFICATION-VALVES							O
O	O		P	XXX	SPECIFICATION-PIPING								O
O	O		J	XXX	SPECIFICATION-INSTRUMENTS									O
O	O		E	XXX	SPECIFICATION-CABLE TRAYS AND FITTINGS										O
O	O		E	XXX	SPECIFICATION-CABLES											O

* Part of the numbers are replaced with “X” for confidentiality.

,

Table A7. Engineering (FBS)-to-Construction (CP/type) relationships.

BLDG		Sys.	Predecessor			Successor *
RCB	AB		OBS	FBS *	Description	CX	CX	SX	PX	EX	EX	EX	JX	MX	MX
O	O		P	XXX	GENERAL ARRANGEMENT DRAWING
O	O	O	All	XXX	SYSTEM DESIGN CRITERIA
X	O	O	M	XXX	P&ID
O	X	O	N	XXX	P&ID
O	O		P	XXX	PIPING STRESS ANALYSIS
O	O		P	XXX	PIPING ISO DRAWING				O					O	O
O	O		P	XXX	PIPING SUPPORT DRAWING				O					O	O
O	O		C	XXX	STRUCTURAL ANALYSIS
O	O		C	XXX	STRUCTURAL DRAWING	O	O
O	O		C	XXX	STRUCTURE DETAIL DRAWING	O	O
O	X		C	XXX	LINER PLATE CALCULATION
O	X		C	XXX	LINER PLATE DRAWING	O	O	O
O	O	O	E	XXX	SINGLE LINE DIAGRAM
O	O		E	XXX	CABLE TRAY DRAWING				O	O		O			O
O	O		E	XXX	CONDUIT PLAN DRAWING	O					O	O
O	O	O	J	XXX	C&ID
O	O		J	XXX	INSTRUMENT LOCATION DRAWING								O
O	O		C	XXX	SPECIFICATION-REINFORCING STEEL BAR	O	O
O	O		C	XXX	SPECIFICATION-READY MIXED CONCRETE	O	O
O	X		C	XXX	SPECIFICATION-CONTAINMENT LINER PLATES			O
X	O		M	XXX	SPECIFICATION-PUMPS										O
O	X		M	XXX	SPECIFICATION-CRANE										O
O	X		M	NSSS	SPECIFICATION-NSSS EQUIPMENT									O
X	O		P	XXX	SPECIFICATION-VALVES				O
O	O		P	XXX	SPECIFICATION-PIPING				O
O	O		J	XXX	SPECIFICATION-INSTRUMENTS								O
O	O		E	XXX	SPECIFICATION-CABLE TRAYS AND FITTINGS					O	O
O	O		E	XXX	SPECIFICATION-CABLES							O

* Part of the numbers are replaced with “X” for confidentiality.

and

Table A8. Procurement (PO)-to-Construction (CP/type) relationships.

BLDG		Sys.	Predecessor			Successor *
RCB	AB		OBS	FBS *	DESCRIPTION	CX	CX	SX	PX	EX	EX	EX	JX	MX	MX
O	O		C	XXX	SPECIFICATION-REINFORCING STEEL BAR	O	O
O	O		C	XXX	SPECIFICATION-READY MIXED CONCRETE	O	O
O	X		C	XXX	SPECIFICATION-CONTAINMENT LINER PLATES			O
X	O		M	XXX	SPECIFICATION-PUMPS										O
O	X		M	XXX	SPECIFICATION-CRANE										O
O	X		M	NSSS	SPECIFICATION-NSSS EQUIPMENT									O
X	O		P	XXX	SPECIFICATION-VALVES				O
O	O		P	XXX	SPECIFICATION-PIPING				O
O	O		J	XXX	SPECIFICATION-INSTRUMENTS								O
O	O		E	XXX	SPECIFICATION-CABLE TRAYS AND FITTINGS					O	O
O	O		E	XXX	SPECIFICATION-CABLES							O

* Part of the numbers are replaced with “X” for confidentiality.

) identified logical linkages among engineering (FBS), procurement (PO), and construction (CP) activity types.

Estimate Activity Durations

Activity durations were determined based on the following: (i) reference values from Korean NPP project schedules; (ii) adjustments reflecting the scaling factors of the hypothetical model (Table 1); and (iii) the simplified activity flow (Figure A8).

• Engineering activities: PI (3 months), CI (2 months), and specification issue for ITT (2 months).
• Procurement activities: ITT/PO/VD (2 months), R&A (1 month), and fabrication & site delivery (1–12 months depending on material or equipment type).
• Construction activities: durations were assigned according to area, level, or type of work, and adjusted for scaling ratios, reflecting typical field productivity in Korean NPP projects.

This standardized approach ensured comparability between the traditional baseline and AWP-based schedules.

Development Schedule

The traditional baseline schedule was developed using a CPM methodology in Microsoft Excel and Primavera P6. Logical networks were constructed for each Level-4 WBS element and linked according to Figure A8 and * Part of the numbers are replaced with “X” for confidentiality.

Table A6, Table A7 and Table A8 to reflect typical NPP construction flows. Before import into P6, an Excel Gantt chart was prepared and reviewed to validate linkages, confirm EPC sequencing, and resolve inconsistencies. After validation, activity sequences and durations were imported into P6 to generate the baseline traditional project schedule.

3.3. AWP-Based Schedule Development

3.3.1. WBS Creation

WBS Structure

The AWP-based WBS was configured with reference to established AWP WBS frameworks [24] and aligned with CWA segmentation to support the development of EWPs, PWPs, and CWPs (Figure A2). The activity-development process followed the traditional baseline and was further refined into a CWA-based organization to reflect AWP principles.

Define Work Packages

Four alternative packaging schemes were developed from the defined WBS (Figure A3, Figure A4, Figure A5 and Figure A6) to test how traditional engineering and procurement processes could be structured into AWP-compatible units (

Table 2. Structural comparison of methods 1 to 4.

Feature	Method 1	Method 2	Method 3	Method 4
EWP grouping	FBS (horizontal)	Discipline (horizontal + vertical)	Disc. + Phase (PI/CI) (vertical)	Discipline (horizontal + vertical)
PWP grouping	Equipment type (horizontal)	Equipment type (horizontal)	Equipment by phase (vertical)	Discipline (horizontal + vertical)

). The four alternatives are characterized as follows:

• Method 1 (AWP-M1): Engineering by functional deliverable (FBS); procurement by equipment type.
• Method 2 (AWP-M2): Engineering by discipline; procurement by equipment type.
• Method 3 (AWP-M3): Engineering by discipline and by issue phase (PI/CI); procurement by discipline and by procurement phase.
• Method 4 (AWP-M4): Engineering and procurement by discipline.

Based on applicability to CWA-based scheduling, corroborated by the assessment criteria and scores in Appendix A.1, methods 1 and 2 were selected for detailed schedule development. The rationale for selection and their comparative performance are reported in Section 4.

3.3.2. Schedule Development

Define Activities

The activity structure followed the same framework and standard relationships as the traditional baseline (Figure A8). Activities were then subdivided by CWA under the AWP-based WBS, enabling independent planning and tracking by work area and ensuring alignment among EWPs, PWPs, and CWPs while preserving comparability.

Sequence Activities

Sequencing was derived by partitioning the traditional activity flows by CWA and applying the POC. Although traditional Korean NPP schedules did not explicitly use the terms “CWA” and “POC,” equivalent concepts are partly embedded in segmentation practices (e.g., design areas, construction joints/sequences). In this study, the same CWA segmentation and POC was applied to both the traditional and AWP-based schedules for construction.

Estimate Activity Durations

Activity durations were adopted from the traditional baseline to ensure consistency. The same 30-day lead time was applied between engineering/procurement completion and the start of corresponding construction activities. This approach isolates structural effects of work packaging and sequencing.

Development Schedule

AWP-based schedules were developed directly in Primavera P6 using the CWA-based WBS and activity definitions. Common project milestones were maintained across models to enable direct comparison of schedule performance and characteristics.

3.4. Schedule Impact Adoption

3.4.1. Construction Productivity Effect

A 25% construction-productivity improvement factor was applied to AWP-M1 and AWP-M2 to evaluate the potential benefits of AWP-based delivery. Prior studies [7,9,10] report field-productivity gains of 20–25%, attributed to improved workface planning (reduced idle time). Accordingly, durations of construction activities in Primavera P6 were reduced by 25% in the AWP-based schedules to isolate AWP-driven productivity effects. In this preliminary scenario, the 25% uplift is used as a representative value synthesized from industry case studies [9] and is not a trade-specific estimate. The purpose is to provide an intuitive, reproducible comparison of schedule behavior between traditional and AWP configurations without over-fitting to sparse or non-public domestic datasets.

In addition, to reflect the AWP practice that Engineering Work Packages are issued at least 90 days before field execution, a 90-day finish-to-start lag was inserted between IEWP and the start of corresponding construction; the same minimum lead time was enforced between completion of procurement and construction start. This lead-time setting ensured engineering completion preceded construction with sufficient margin for IWP development, improving sequencing control and compliance with workface-planning protocols.

Together, these modifications enabled a structured comparison between the traditional baseline and the AWP-based models under uplifted productivity assumptions and explicit readiness constraints.

3.4.2. Workface Planning Effect

In field operations, the workface planner verified information completeness for IWPs and monitored whether planned construction could proceed as scheduled. When information was incomplete or constraints remained near a planned start, the planner flexibly adjusted sequencing using a prepared backlog of IWPs to prevent delays.

Assuming that construction began only after information completeness was ensured, a delay scenario was designed and implemented to examine the schedule impact of engineering delays and to evaluate AWP-based WFP mitigation (In the simplified baseline schedule, procurement activities were placed explicitly between engineering deliverables and construction tasks (Figure A8). Any procurement delay therefore propagates deterministically to site installation. For this reason, the simulation scope was limited to engineering delays, as their downstream impact inherently included procurement slippage.). The procedure (Figure 3) comprised the following steps.

• Selection of delay-causing engineering activities. As summarized in

  Table A9. Delay causing engineering activities for each schedule *.

  CWA	Disc	FBS	Activity ID
  			Traditional	AWP-M1		AWP-M2
  				A	B	A	B
  1: RCB	C	1XX	3XXC1XXPI	CWP01-EWP10-3XX1C1XXPI	I-EWP 10	CWP01-EWP9-3XX1C1XXPI	I-EWP 9
  	E	1XX	3XXE1XXPI	CWP03-EWP16-3XX1E1XXPI	I-EWP 16	CWP03-EWP13-3XX1E1XXPI	I-EWP 13
  	J	1XX	3XX1XXPI	CWP06-EWP24-3XX1J1XXPI	I-EWP 24	CWP06-EWP21-3XX1J1XXPI	I-EWP 21
  	P	1XX	3XXP1XXPI	CWP07-EWP27-3XX1P1XXPI	I-EWP 27	CWP03-EWP25-3XX1P1XXPI	I-EWP 25
  2: AB	C	1XX	3XXC1XXPI	CWP09-EWP41-3XX2C1XXPI	I-EWP 41	CWP09-EWP31-3XX2C1XXPI	I-EWP 31
  	E	1XX	3XXE1XXPI	CWP00-EWP46-3XX2E1XXPI	I-EWP 46	CWP00-EWP35-3XX2E1XXPI	I-EWP 35
  	J	1XX	3XXJ1XXPI	CWP12-EWP52-3XX2J1XXPI	I-EWP 52	CWP12-EWP40-3XX2J1XXPI	I-EWP 40
  	P	1XX	3XXP1XXPI	CWP10-EWP57-3XX2P1XXPI	I-EWP 57	CWP10-EWP44-3XX2P1XXPI	I-EWP 44

  * Part of the numbers are replaced with “X” for confidentiality.

  , delay drivers were selected from the Traditional, AWP-M1, and AWP-M2 schedules by choosing, for each discipline, the earliest drawing-related engineering activity that could affect downstream construction. Multi-discipline deliverables with broad successor impacts were excluded to enable clear attribution of WFP effects. Target structures were grouped as RCB and AB; for the AB, a conservative choice selected the earliest engineering activity in CWA2 to ensure a direct, measurable impact on subsequent construction.
• Insertion of delaying dummy activity. A Delaying Dummy Activity was inserted as a predecessor to the selected construction tasks to represent upstream engineering delay.
• Increment of dummy duration. The dummy duration was increased in monthly steps using.

N = n + a × i

(1)

where

• Initial delay: n = 0;
• Increment step: a = 30 days;
• Iteration step: i = i + 1;
• Maximum delay duration: N ≤ 360 days.

4.

Schedule delay analysis. For each increment, the schedule was recalculated in Primavera P6, reflecting resulting changes in milestone dates.

5.

Evaluation of milestone impacts. Shifts were confirmed for First Concrete, Set Reactor Vessel (RV), and Construction Finish.

6.

WFP mitigation scenarios. Two scenarios were applied independently to isolate individual effects:

• Scenario ①—Lead-time reduction by construction readiness. Consistent with CII IR 272-2 guidance [7], IWPs were assumed to be prepared 3 months (2 months development and 1 month of lead time) in advance. Upon detecting delay, up to 30 days of this 3-month window were repurposed as schedule float without compromising readiness.
• Scenario ②—Risk-driven re-sequencing via backlog utilization. Successor activities were re-sequenced to maintain workflow continuity, allowing non-impacted areas to proceed in parallel using pre-issued backlogs; the logic is summarized in

  Table A10. Re-sequencing steps.

  Steps	Relationships
  Baseline	Activities A → B → C follow a standard Finish-to-Start (FS) logic.
  Delay Occurrence	A delay in activity A causes a shift in B and C.
  Re-Sequencing	While activity A is delayed, activity B is advanced independently by resequencing the relationship, reducing the cumulative schedule impact.

  .

• Initially, Scenario ① was intended to run with concurrent re-sequencing; however, extensive successor dependencies around the delayed engineering tasks made practical re-sequencing infeasible. Consequently, Scenario ② was executed separately by substituting the delayed engineering activity with its IEWP counterpart, enabling a distinct assessment of the re-sequencing strategy and clearer attribution of milestone recovery.

7.

Schedule update and recording. After each schedule delay/mitigation analysis, the schedule was updated and milestone impacts were recorded.

8.

Iteration and termination. Steps 3–7 were repeated for all increments until N = 360 days, at which point the simulation terminated.

3.5. Comparative Analysis

This section presents a comparative analysis of the traditional baseline and the two AWP-based schedules (AWP-M1 and AWP-M2). The objective was to evaluate, in a controlled manner, how differences in work packaging, WBS structure, activity sequencing, assumed productivity uplift, and WFP mitigation affected schedule performance and resilience under engineering-delay conditions. To ensure consistency, all models shared the same overall project scope and milestones, top-level WBS hierarchy, activity taxonomy, and base sequencing rules. The analysis evaluated (i) WBS structure—level of detail, hierarchy alignment, and integration with CWA and POC; (ii) activity and deliverable breakdown—counts, types, and package distribution; (iii) schedule durations—overall and by phase; (iv) sensitivity to engineering delays—shifts in construction finish; and (v) mitigation performance under delay scenarios—effects of WFP actions and recovery patterns relative to the traditional model. The quantitative results in Section 4 demonstrate the schedule impacts of AWP-based methods, outline advantages and limitations of each approach, and indicate their applicability to Korean NPP project environments.

4. Results: Illustrative Application and Discussion

4.1. Traditional Schedule Development

4.1.1. WBS Creation

The traditional WBS was developed in line with typical Korean NPP practices [24]. The structure supported the study objectives by enabling comparison with AWP-based WBS models and by providing a basis for schedule development. It followed a discipline-driven, phase-based hierarchy segmented into EPC phases and was optimized for progress measurement and cost management. For consistency across phases, activities were grouped by discipline using an Organizational Breakdown Structure (OBS), even though actual projects often distinguish engineering functions from contract packaging (e.g., NSSS vs. BOP, or construction by package vs. by discipline).

The WBS and corresponding activities were first assembled in excel to reflect major deliverables (e.g., reinforcing steel bars, structural drawings) and included attribute fields to enable direct import into Primavera P6. A summary of the total number of activities and deliverables by WBS level is provided in

Table 3. Traditional—Summary of activity and deliverable quantities by WBS.

WBS Level			Quantity of Activities					ENG Deliverable
Lev.2	Lev.3	Lev.4	AE/BOP	NSSS/RCB	Common	Interface	Total
Milestone/Dummy			-	-	3	-	3	-
ENG	6	28	24	31	10	98	163	269
PRO	5	11	0	9	24	59	92	-
CON	5	9	89	46	0	124	259	-
4	16	48	113	86	37	281	517	269

(details in

Table A11. Traditional—Activity and deliverable quantities by WBS.

WBS Level			Quantity of Activities					ENG Deliverable
Lev.2	Lev.3	Lev.4 Amount	AE/BOP	NSSS/RCB	Common	Interface	Total
Milestone/Dummy			-	-	3	-	3	-
ENG	C	8	5	10	2	42	59	91
	E	5	4	4	4	16	28	52
	J	3	4	4	1	3	12	29
	M	5	2	4	1	20	27	86
	P	6	7	7	2	17	33	7
	X	1	2	2	0	0	4	4
	Sum	28	24	31	10	98	163	269
PRO	C	3	-	3	6	17	26	-
	E	2	-	-	6	11	17	-
	J	1	-	-	3	4	7	-
	M	3	-	6	3	13	22	-
	P	2	-	-	6	14	20	-
	Sum	11	0	9	24	59	92	-
CON	C(A)	3	36	23	-	50	109	-
	E	2	17	6	-	21	44	-
	J	1	4	1	-	4	9	-
	M	2	16	11	-	24	51	-
	P	1	16	5	-	25	46	-
	Sum	9	89	46	0	124	259	-
4	16	48	113	86	37	281	517	269

).

4.1.2. Schedule Development

An Excel Gantt chart was prepared from the relationship mapping in * Part of the numbers are replaced with “X” for confidentiality.

Table A6, Table A7 and Table A8 to verify activity linkages prior to development in Primavera P6. After validation, the activities and relationships were imported into P6 to generate the traditional schedule; the resulting critical activities are shown in Figure A12. Key project milestones are summarized in

Table 4. Key project milestones.

Milestone	Scheduled Date	From Project Start
Project Start (D)	2026.01.01	D + 0 month
First Concrete	2027.11.05	D + 22 month
Set Reactor Vessel (RV)	2028.12.09	D + 35 month
Construction Completion	2029.06.11	D + 42 month

, yielding a total project duration of approximately 42 months from start to finish across engineering, procurement, and construction phases. The quantity and types of logic relationships applied in the P6 schedule are summarized in

Table 5. Traditional-P6 relationships.

Predecessor–Successor	Relationship Type	Amount
Milestone ⇒ CON	FS	2
ENG ⇒ ENG	FS	102
ENG ⇒ PRO	FS	22
ENG ⇒ CON	FS	174
PRO ⇒ PRO	FS	22
PRO ⇒ CON	FS	74
CON ⇒ CON	SS	8
	FS	178
	FF	7
		589

, with a total of 589 predecessor–successor links.

4.2. AWP-Based Schedule Development

4.2.1. WBS Creation

WBS Structure

In developing the AWP-based WBS, Figure A2 was referred to as the foundational guideline. As shown in Figure 4, the hierarchical structure was reorganized to align with the AWP framework, emphasizing physical construction zones (CWAs), work packaging logic, and clear alignment with the EWPs, PWPs, and CWPs.

In contrast to the traditional schedule, where the breakdown was driven by functional disciplines or contract packages, the AWP-based WBS was configured as follows:

• Level 1: Project;
• Level 2: CWA;
• Level 3: Phase (Engineering, Procurement, Construction);
• Level 4: Discipline (OBS);
• Level 5: Work Package;
• Level 6: FBS, Equipment type, Construction type.

This configuration maintained compatibility with project-management practice while aligning with AWP. Selecting CWA at Level 2 reflected the central role of physical workface zones in defining IWPs and CWPs. Clear phase delineation at level 3 enabled distinct control of engineering, procurement, and construction workflows. An OBS-based level 4 was applied uniformly to harmonize the hierarchy across phases and to improve responsibility clarity and traceability. Although real projects often group procurement by NSSS vs. BOP or organize construction by package/subcontractor, such contract-driven groupings were intentionally avoided here to enhance consistency and comparability across WBS levels. In addition, whereas traditional schedules typically use POs as the primary procurement unit, the AWP WBS decomposed these elements under each CWA; consequently, PWPs should be regrouped and redefined in line with procurement/contracting strategies that reflect this breakdown.

Define Work Packages

To establish an effective AWP packaging framework, four alternatives were evaluated with an assessment matrix (Appendix A.1) against six criteria—Scope Definition, WBS Compatibility, Schedule Unit, Compatibility with Schedule Hierarchy, Constructability, and Progress Measurement—for both EWP and PWP.

• Method 1 (AWP-M1). Scored 4–5 across all criteria and was the only scheme to receive a 5 for both “EWP as a Schedule Unit” and “Progress Measurement by EWP.” The FBS-based EWP boundary forms self-contained schedule units with minimal cross-coupling, and its deliverable register enables direct, auditable progress measurement. Equipment-type PWPs nest cleanly under CWA/CWP and align with procurement/fabrication streams, sustaining constructability with routine materials planning and measurable progress.
• Method 2 (AWP-M2). Also scored 4–5 across all criteria and was the only scheme to receive a 5 for “EWP Compatibility with Schedule Hierarchy.” Discipline-based EWPs map directly to the OBS layer within the WBS, slotting into the schedule hierarchy with supplemental links. PWPs are defined the same as in Method 1.
• Method 3 (AWP-M3). Predominantly 3–4. The EWP scored 3 on “Scope Definition, Schedule Unit, and Compatibility with Schedule Hierarchy” because splitting by issue phase (PI/CI) divides a single discipline’s scope over time, creating relationship logics for overlapping activities of phase functions of work, and an expanding number of packages as phases increase. The PWP likewise scored 3 on the same items (and associated constructability/progress implications) since discipline-plus-procurement-phase coupling conflicts with PO lifecycles (composition and process), demands supplementary activity relationships and field-requested-date plans, and require management plans per equipment-level traceability.
• Method 4 (AWP-M4). Ranged 3–5. The EWP have same scores with method 2, but the PWP scored 3 on “Schedule Unit, Constructability, and Progress Measurement” because discipline-based PWPs conflicts with PO lifecycles (composition), demands supplementary activity relationships and field-requested-date plans, and require management plans per equipment-level traceability.

A recurring concern with PWPs was granularity. Decomposition by CWA/CWP further multiplies the amount of PWPs; moreover, site logistics and field-requested dates make pure WP-level materials control burdensome without sub-WP checks. Accordingly, in this study, we retain PWP grouping for WBS/control purposes but do not introduce separate “PWP issuance” activities; procurement is anchored in the schedule by fabrication and delivery nodes that already tie to engineering readiness and field need dates. Considering the above results and the scores in

Table 6. Score summary of the packaging assessment matrix.

Method	Total Score	Converted Score
Method 1	56/60	93.6%
Method 2	55/60	92.0%
Method 3	41/60	69.2%
Method 4	50/60	83.2%

, Methods 1 and 2 were advanced to detailed schedule development, while Methods 3 and 4 were excluded due to phase-split inefficiencies (M3) and discipline-based PWP control/logistics drawbacks (M4).

AWP Method 1 (AWP-M1)

The WBS and corresponding activities for AWP-M1 were first developed in Excel to reflect major deliverables (e.g., reinforcing steel bars, structural drawings) and included attribute fields for direct import into Primavera P6. Table 7 (details in

Table A12. AWP-M1-Activity and deliverable quantities by WBS.

WBS Level				Quantity (Activities/IEWP/ENG-IF)				ENG Deliverable
Lev.2 CWA	Lev.3 Phase	Lev.4 OBS	Lev.5 WP	AE/BOP	NSSS/RCB	Common	Total
Milestone/Dummy				-	-	3	3	-
0	ENG	E	2	2 (1/1/0)	-	3 (2/1/0)	5 (3/2/0)	3
		J	1	2 (2/0/0)	3 (2/1/0)	-	5 (4/1/0)	4
		M	2	2 (2/0/0)	5 (3/2/0)	-	7 (5/2/0)	5
		X	1	2 (2/0/0)	3 (2/1/0)	-	5 (4/1/0)	4
		Sum	6	8 (7/1/0)	11 (7/4/0)	3 (2/1/0)	22 (16/6/0)	16
	PRO	E	1	3	-	-	3	-
		M	1	-	3	-	3	-
		Sum	2	3	3	-	6	-
	CON	E	1	1	-	-	1	-
		Sum	1	1	-	-	1	-
	Sum		9	12	14	3	29	16
1	ENG	C	8	-	20 (12/8/-)	-	20 (12/8/-)	66
		E	9	-	33 (16/9/8)	-	33 (16/9/8)	100
		J	2	-	5 (3/2/-)	-	5 (3/2/-)	15
		M	2	-	4 (2/2/-)	-	4 (2/2/-)	93
		P	10	-	31 (15/10/6)	-	31 (15/10/6)	2
		Sum	31	-	93 (48/31/14)	-	93 (48/31/14)	276
	PRO	C	3	-	9	-	9	-
		E	2	-	6	-	6	-
		J	1	-	3	-	3	-
		M	2	-	6	-	6	-
		P	2	-	6	-	6	-
		Sum	10	-	30	-	30	-
	CON	A	2	-	23	-	23	-
		E	2	-	6	-	6	-
		J	1	-	1	-	1	-
		M	2	-	11	-	11	-
		P	1	-	5	-	5	-
		Sum	8	-	46	-	46	-
	Sum		49	-	169	-	169	276
2	ENG	C	5	12 (7/5/-)	-	-	12 (7/5/-)	27
		E	8	31 (15/8/8)	-	-	31 (15/8/8)	71
		J	2	5 (3/2/-)	-	-	5 (3/2/-)	11
		M	1	2 (1/1/-)	-	-	2 (1/1/-)	52
		P	8	26 (12/8/6)	-	-	26 (12/8/6)	1
		Sum	24	76 (38/24/14)	-	-	76 (38/24/14)	162
	PRO	C	2	6	-	-	6	-
		E	1	3	-	-	3	-
		J	1	3	-	-	3	-
		M	1	3	-	-	3	-
		P	2	6	-	-	6	-
		Sum	7	21	-	-	21	-
	CON	A	1	9	-	-	9	-
		E	1	4	-	-	4	-
		J	1	4	-	-	4	-
		M	1	4	-	-	4	-
		P	1	4	-	-	4	-
		Sum	5	25	-	-	25	-
	Sum		36	122	-	-	122	162
3	Same as CWA 2
	Sum		36	122	-	-	122	162
4	Same as CWA 2
	Sum		36	122	-	-	122	162
5	Same as CWA 2
	Sum		36	122	-	-	122	162
Total			202	500	183	6	689	940

) summarizes activity quantities by WBS level, including IEWP and ENG-IF items (definitions provided in Section 4.2.2). The AWP-M1 WBS was then implemented in P6 following the standard logical structure in Figure A3.

Table 7. AWP-M1-Summary of activity and deliverable quantities by WBS.

WBS Level				Quantity (Activities/IEWP/ENG-IF)				ENG Deliverable
Lev.2 CWA	Lev.3 Phase	Lev.4 OBS	Lev.5 WP	AE/BOP	NSSS/RCB	Common	Total
Milestone/Dummy				-	-	3	3	-
0	ENG	4	6	8 (7/1/0)	11 (7/4/0)	3 (2/1/0)	22 (16/6/0)	16
	PRO	2	2	3	3	-	6	-
	CON	1	1	1	-	-	1	-
	Sum		9	12	14	3	29	16
1	ENG	5	31	-	93 (48/31/14)	-	93 (48/31/14)	276
	PRO	5	10	-	30	-	30	-
	CON	5	8	-	46	-	46	-
	Sum		49	-	169	-	169	276
2, 3, 4, 5 each	ENG	5	24	76 (38/24/14)	-	-	76 (38/24/14)	162
	PRO	5	7	21	-	-	21	-
	CON	5	5	25	-	-	25	-
	Sum		36	122	-	-	122	162
Total			202	500	183	6	689	940

Table 7. AWP-M1-Summary of activity and deliverable quantities by WBS.

WBS Level				Quantity (Activities/IEWP/ENG-IF)				ENG Deliverable
Lev.2 CWA	Lev.3 Phase	Lev.4 OBS	Lev.5 WP	AE/BOP	NSSS/RCB	Common	Total
Milestone/Dummy				-	-	3	3	-
0	ENG	4	6	8 (7/1/0)	11 (7/4/0)	3 (2/1/0)	22 (16/6/0)	16
	PRO	2	2	3	3	-	6	-
	CON	1	1	1	-	-	1	-
	Sum		9	12	14	3	29	16
1	ENG	5	31	-	93 (48/31/14)	-	93 (48/31/14)	276
	PRO	5	10	-	30	-	30	-
	CON	5	8	-	46	-	46	-
	Sum		49	-	169	-	169	276
2, 3, 4, 5 each	ENG	5	24	76 (38/24/14)	-	-	76 (38/24/14)	162
	PRO	5	7	21	-	-	21	-
	CON	5	5	25	-	-	25	-
	Sum		36	122	-	-	122	162
Total			202	500	183	6	689	940

AWP Method 2 (AWP-M2)

The WBS and activities for AWP-M2 were likewise developed in Excel with import attributes for Primavera P6. In contrast to AWP-M1, AWP-M2 bundled multiple deliverables into fewer work packages, reducing package counts and simplifying coordination, while it may increase the complexity of tracking individual deliverable readiness or constraint resolution due to the higher level of aggregation within each package. This approach reflects field practice in which engineering drawings are often issued and reviewed in batches by area or system.

Table 8 presents the quantitative summary for AWP-M2 (including IEWP and ENG-IF counts; Section 4.2.2), showing reductions in both activities and packages relative to AWP-M1—primarily due to integration of CWP-related engineering work. The resulting AWP-M2 WBS was configured in P6 based on the logical structure in Figure A4.

Table 8. AWP-M2-Summary of activity and deliverable quantities by WBS.

WBS Level				Quantity (Activities/IEWP/ENG-IF)				ENG Deliverable
Lev.2 CWA	Lev.3 Phase	Lev.4 OBS	Lev.5 WP	AE/BOP	NSSS/RCB	Common	Total
Milestone/Dummy				-	-	3	3	-
0	ENG	4	6	8 (1/1/0)	11 (7/4/0)	3 (2/1/0)	22 (16/6/0)	16
	PRO	2	2	3	3	-	6	-
	CON	1	1	1	-	-	1	-
	Sum		9	12	14	3	29	16
1	ENG	5	24	-	86 (48/24/14)	-	86 (48/24/14)	276
	PRO	5	10	-	30	-	30	-
	CON	5	8	-	46	-	46	-
	Sum		42	-	162	-	162	276
2, 3, 4, 5 each	ENG	5	16	68 (38/16/14)	-	-	68 (38/16/14)	162
	PRO	5	7	21	-	-	21	-
	CON	5	5	25	-	-	25	-
	Sum		28	114	-	-	114	162
Total			163	468	176	6	650	940

Table 8. AWP-M2-Summary of activity and deliverable quantities by WBS.

WBS Level				Quantity (Activities/IEWP/ENG-IF)				ENG Deliverable
Lev.2 CWA	Lev.3 Phase	Lev.4 OBS	Lev.5 WP	AE/BOP	NSSS/RCB	Common	Total
Milestone/Dummy				-	-	3	3	-
0	ENG	4	6	8 (1/1/0)	11 (7/4/0)	3 (2/1/0)	22 (16/6/0)	16
	PRO	2	2	3	3	-	6	-
	CON	1	1	1	-	-	1	-
	Sum		9	12	14	3	29	16
1	ENG	5	24	-	86 (48/24/14)	-	86 (48/24/14)	276
	PRO	5	10	-	30	-	30	-
	CON	5	8	-	46	-	46	-
	Sum		42	-	162	-	162	276
2, 3, 4, 5 each	ENG	5	16	68 (38/16/14)	-	-	68 (38/16/14)	162
	PRO	5	7	21	-	-	21	-
	CON	5	5	25	-	-	25	-
	Sum		28	114	-	-	114	162
Total			163	468	176	6	650	940

4.2.2. Schedule Development

The AWP-based schedules were developed using the same foundational logic and standardized sequences as the traditional baseline. Sequencing and activity relationships followed the workflows in Figure A3 and Figure A4 and the logic map in Table A6, Table A7 and Table A8. All activities were defined during WBS development in Section 4.2.1.

However, two critical implementation considerations unique to AWP were identified and incorporated during scheduling:

Management of Duplicate Engineering Deliverables

In procurement planning, overlapping information across multiple PWPs (due to CWP breakdowns) is generally manageable—particularly for physical items such as tags and bulk materials, which remain consistent and identifiable.

However, in engineering, the same deliverables (e.g., drawings, specifications) may be required by multiple CWPs within a single CWA. As the number of CWPs increases or their scopes narrow, the same engineering deliverable can become logically assigned to several different CWPs. In AWP, these are treated as separate entities and assigned different codes. While this is acceptable in a database-driven management system, it introduces complexity when developing and managing the schedule in tools like Primavera P6, which require clear logic relationships.

To address this, Start Interface (SI) and Finish Interface (FI) activities—collectively referred to as ENG-IF—were introduced. These act as centralized control points for shared deliverables across CWPs. Each CWP that references a shared deliverable establishes logical relationships to these ENG-IF activities:

• SI captures the earliest start among all referencing activities.
• FI captures the latest finish among them.

This ensured that predecessor/successor logic, especially for fabrication and fieldwork, depended on the global start/finish of the shared deliverable rather than on any single instance. The configuration is illustrated in Figure A9, and Figure A10 shows how a delay to one EWP shifts the global FI and downstream successors.

Issuance of Engineering Work Packages (IEWP)

In traditional EPC scheduling, engineering, procurement, and construction activities are often managed independently and interface dates are exchanged via designated interface activities. These interface activities are often contractually fixed and rigid.

To align with AWP principles, IEWP activities were introduced to replace interface activities between engineering and construction. These represent the final issuance of engineering deliverables required for CWP-level field execution and serve as the formal control points between design and construction.

While interface activities may still be relevant between engineering and procurement (e.g., for vendor data or specification transfer), the engineering-to-construction relationship is now captured via these IEWP activities.

Procurement activities, in this study, were not further decomposed into physical PWPs. Although PWP-level scheduling could be considered, it was deemed unnecessary since fabrication activities already serve as effective scheduling anchors with direct logical links to engineering deliverables. Thus, schedule management remains both simplified and meaningful.

All construction deliverables required by a CWP are logically connected to their corresponding IEWP activities, ensuring visibility into readiness and schedule integration. Figure A11 shows the added IEWP activity and relationship within EWP16.

Following the application of the WBS and logic development principles described above, the scheduling results for AWP-M1 and AWP-M2 were compiled and summarized. These outcomes include the total quantity of activities, work packages, and engineering deliverables generated, as well as the quantity and types of logical relationships developed in Primavera P6.

Table 7 and Table 8 present the quantitative outcomes of activity and deliverable generation for AWP-M1 and AWP-M2, respectively. AWP-M1 resulted in a total of 202 work packages, 689 activities including IEWPs and ENG-IF nodes, and 940 engineering deliverables. AWP-M2 generated 163 work packages, 650 activities including IEWPs and ENG-IF nodes, and 940 engineering deliverables.

Logical relationship mappings used in P6 are summarized in

Table 9. AWP-M1-P6 relationships.

Predecessor–Successor	Relationship Type	Amount
Milestone ⇒ CON	FS	2
ENG ⇒ ENG	FS	144
ENG ⇒ IEWP	FS	136
ENG ⇒ ENG-IF	FS	123
ENG ⇒ PRO	FS	40
IEWP ⇒ CON	FS	96
ENG-IF ⇒ ENG	FS	123
ENG-IF ⇒ ENG-IF	FS	24
ENG-IF ⇒ CON	FS	57
PRO ⇒ PRO	FS	80
PRO ⇒ CON	FS	40
CON ⇒ CON	SS	8
	FS	187
	FF	19
		1079

and

Table 10. AWP-M2-P6 relationships.

Predecessor–Successor	Relationship Type	Amount
Milestone ⇒ CON	FS	2
ENG ⇒ ENG	FS	144
ENG ⇒ IEWP	FS	136
ENG ⇒ ENG-IF	FS	123
ENG ⇒ PRO	FS	40
IEWP ⇒ CON	FS	70
ENG-IF ⇒ ENG	FS	123
ENG-IF ⇒ ENG-IF	FS	24
ENG-IF ⇒ CON	FS	57
PRO ⇒ PRO	FS	80
PRO ⇒ CON	FS	40
CON ⇒ CON	SS	8
	FS	187
	FF	19
		1053

. AWP-M1 included 1079 relationships and AWP-M2 slightly reduced the total to 1055, primarily due to a decreased number of IEWP to CON dependencies.

Developed Schedule

The final AWP schedules critical path activities are visualized in Figure A13 and Figure A14. AWP-M1 shows more IEWP activities and relationships due to finer packaging granularity, whereas AWP-M2 applies more aggregated packaging with fewer work packages and slightly simpler paths. Both AWP schedules maintained the same key milestone dates as the traditional model (Table 4).

4.3. Schedule Impact Adoption

4.3.1. Construction Productivity Effect

Applying a 25% construction-productivity uplift to AWP-M1 and AWP-M2 advanced the Construction Finish by 120 days. To comply with AWP practice that EWPs are issued at least 90 days before field execution, a 90-day IEWP → CWP finish-to-start lag was enforced; this offset part of the gain, yielding a net improvement of 60 days versus the baseline. This resulted from a 120-day reduction achieved by the 25% productivity uplift, partly offset by a 60-day increase associated with the readiness lag for IWP development under AWP practice, yielding a net saving of 60 days. The resulting dates and variances are summarized in

Table 11. Schedule Outcomes under Construction Productivity Adjustment.

Scenario	Finish Date	Schedule Variance	Remarks
Baseline Schedule (No Adjustment) (ⓐ)	11 June 2029	±0 days	AWP-M1, M2 schedule without AWP improvements
25% Productivity Improvement Applied (ⓑ)	11 February 2029	−120 days (ⓑ − ⓐ)	Reflects 25% reduction in CWP durations
ⓑ + 90 day Lag for IEWPto CWP Sequencing (ⓒ)	12 April 2029	−60 days (ⓒ − ⓐ)	Accounts for AWP workface planning requirement

.

4.3.2. Workface Planning Effect

Delay Impact Results in Traditional Schedule

In the Traditional model, the insertion of 360-day delays into eight discipline-specific engineering activities resulted in milestone deferrals ranging from 0 to 270 days, depending on the degree of float and the criticality of each path. As summarized in

Table 12. Traditional schedule engineering activity delay results.

Delay Activity	Finish Date (A)	Variance (A-Baseline)
310C118PI	8 March 2030	270 days
310E145PI	6 January 2030	209 days
310J159PI	11 June 2029	0 days
310P193PI	8 March 2030	270 days
320C118PI	1 December 2029	173 days
320E145PI	30 November 2029	172 days
320J159PI	11 June 2029	0 days
320P193PI	1 December 2029	173 days

, civil I, electric (E), and piping (P) deliverables in both RCB and AB areas exhibited the largest impact. The delay by I&C (J) deliverable did not influence the finish date because of the free float.

Scenario ①: Schedule Recovery by Backlog Float Utilization (AWP-M1, M2)

To simulate AWP-based mitigation (Scenario ①), a 30-day float derived from the standard IWP preparation buffer was applied to each delayed engineering deliverable. The results for AWP-M1 and AWP-M2 are presented in

Table 13. AWP-M1, M2 schedule engineering activity delay/mitigation scenario results.

Delay Activity		360-Day Delay		30-Day Backlog
AWP-M1	AWP-M2	Finish Date (A)	Variance (A-Baseline)	Finish Date (B)	Variance (B-A)
CWP01-EWP10-3101C118PI	CWP01-EWP9-3101C118PI	7 January 2030	270	8 December 2029	−30
CWP03-EWP16-3101E145PI	CWP03-EWP13-3101E145PI	20 November 2029	222	21 October 2029	−30
CWP06-EWP24-3101J159PI	CWP06-EWP21-3101J159PI	12 April 2029	0	12 April 2029	0
CWP07-EWP27-3101P186PI	CWP03-EWP25-3101P193PI	7 January 2030	270	8 December 2029	−30
CWP09-EWP41-3202C118PI	CWP09-EWP31-3202C118PI	15 November 2029	217	16 October 2029	−30
CWP00-EWP46-3202E145PI	CWP00-EWP35-3202E145PI	29 December 2029	261	29 November 2029	−30
CWP12-EWP52-3202J159PI	CWP12-EWP40-3202J159PI	12 April 2029	0	12 April 2029	0
CWP10-EWP57-3202P193PI	CWP10-EWP44-3202P193PI	15 November 2029	217	16 October 2029	−30

, showing consistent delay reductions of 30 days in most activities where mitigation was applicable. Activities with sufficient float (J159PI) showed no delay regardless of the scenario.

Scenario ②: Schedule Recovery by Re-Sequencing of IEWPs Successor Activities (AWP-M1, M2)

Table 14. AWP-M1 schedule I-EWP activity delay/mitigation scenario results.

AWP-M1	360-Day Delay		Re-Sequencing
Delay Activity	Finish Date	Variance (A-Baseline)	Finish Date (B)	Variance (B-Baseline)	Variance (B-A)
I-EWP10	15 October 2029	186	15 October 2029	186	0
I-EWP16	12 April 2029	0	12 April 2029	0	0
I-EWP24	12 April 2029	0	12 April 2029	0	0
I-EWP27	16 May 2029	34	16 May 2029	34	0
I-EWP41	14 August 2029	124	22 July 2029	0	−23
I-EWP46	12 April 2029	0	12 April 2029	0	0
I-EWP52	12 April 2029	0	12 April 2029	0	0
I-EWP57	14 February 2030	170	17 July 2029	96	−74

presents the results of schedule delay and mitigation outcomes for delayed IEWP activities in the AWP-M1 model. The scenario assumes a 360-day engineering delay, with risk-driven re-sequencing applied to assess potential schedule recovery.

I-EWP16, I-EWP24, I-EWP46, and I-EWP52 cases exhibited no schedule impact despite being delayed, due to the presence of sufficient free float within the schedule. These IEWPs were preconfigured with generous float periods, enabling delay absorption without affecting downstream activities.

I-EWP10 and I-EWP27, both located in the RCB zone, experienced delays of 186 days and 34 days, respectively. However, no schedule recovery was achieved through re-sequencing. This is attributable to the absence of alternative activities within the same work type and time window in adjacent CWAs. Due to their location within the critical structural scope of the Reactor Containment Building, schedule flexibility was highly constrained.

In contrast, I-EWP41 and I-EWP57, associated with the Auxiliary Building (AB), CWA2, demonstrated substantial recovery potential. I-EWP41 initially delayed by 124 days achieved a recovery of 23 days, while I-EWP57, initially delayed by 170 days, recovered 74 days. These outcomes were made possible by the availability of identical work types across adjacent CWAs (CWA3, 4, and 5), allowing for effective re-sequencing and float absorption through parallel activity adjustments.

Table 15. AWP-M2 schedule I-EWP activity delay/mitigation scenario results.

AWP-M2	360-Day Delay		Re-Sequencing
Delay Activity	Finish Date	Variance (A-Baseline)	Finish Date (B)	Variance (B-Baseline)	Variance (B-A)
I-EWP9	15 December 2029	247	15 December 2029	247	0
I-EWP13	12 April 2029	0	12 April 2029	0	0
I-EWP21	12 April 2029	0	12 April 2029	0	0
I-EWP25	14 October 2029	185	14 October 2029	185	0
I-EWP31	14 October 2029	185	1 August 2029	111	−74
I-EWP35	12 April 2029	0	12 April 2029	0	0
I-EWP40	12 April 2029	0	12 April 2029	0	0
I-EWP44	29 November 2029	231	16 September 2029	157	−74

presents the results of schedule delay and mitigation outcomes for delayed IEWP activities in the AWP-M2 model with the same scenario assumptions.

I-EWP13, I-EWP21, I-EWP35, and I-EWP40 cases exhibited no schedule impact despite being delayed, due to the presence of sufficient free float within the schedule. These IEWPs were preconfigured with generous float periods, enabling delay absorption without affecting downstream activities.

I-EWP9 and I-EWP25, both located in the RCB zone, experienced delays of 247 days and 185 days, respectively. However, no schedule recovery was achieved through re-sequencing. This is attributable to the absence of alternative activities within the same work type and time window in adjacent CWAs. Due to their location within the critical structural scope of the Reactor Containment Building, schedule flexibility was highly constrained.

In contrast, I-EWP31 and I-EWP44, associated with the Auxiliary Building (AB), CWA2, demonstrated substantial recovery potential. I-EWP31 initially delayed by 185 days, recovered 74 days, resulting in a net delay of 111 days. I-EWP44 initially delayed by 231 days, recovered 74 days, resulting in a net delay of 157 days. These outcomes were made possible by the availability of identical work types across adjacent CWAs (CWA3, 4, and 5), allowing for effective re-sequencing and float absorption through parallel activity adjustments.

4.4. Comparative Analysis

4.4.1. WBS Structure and Work Package Configuration

As summarized in

Table 16. Structural comparison of WBS & WPs.

Attribute	Traditional Schedule *	AWP-M1	AWP-M2	Remark
WBS (or WP) Levels	4	5	5	* Varies on project and purpose
WBS Level definition	1-Project 2-Phase(EPC) 3-Disc/CP 4-FBS/PO/Type	1-Project 2-CWA 3-Phase(EPC) 4-Discipline 5-WP	1-Project 2-CWA 3-Phase(EPC) 4-Discipline 5-WP
WP grouping	Function -ENG: FBS -PRO: PO -CON: CP	CWA + Function -ENG: FBS -PRO: Equip. type -CON: Con. type	CWA + Function -ENG: Disc. -PRO: Equip. type -CON: Con. type
WPs	48	202	163	* Lowest WBS level
ENG deliverables	269	940	940

* In the traditional schedule attribute, values may vary depending on the project characteristics.

, the Traditional schedule model adopts a WBS that reflects the characteristics of each EPC phase. At Level 2 and below, classification criteria varied by phase—e.g., contract packages in procurement, functional disciplines in engineering, or construction packages and types in field execution. This approach offers operational efficiency for each domain but lacks consistency across the entire schedule. Consequently, cross-phase reporting often requires manual aggregation or selective filtering, which introduces complexity in project controls.

In contrast, both AWP-M1 and AWP-M2 adopt a standardized and hierarchical WBS structure aligned with AWP principles. This configuration ensures consistent traceability and management across all scopes. The adoption of CWAs at Level 2 reflects the physical segmentation of the construction site, and identical breakdown logic is applied throughout the EPC domains and are further grouped by discipline, equipment type, or construction type depending on the phase.

However, such standardized structure can lead to increased fragmentation. Elements previously consolidated in the traditional schedule (e.g., common engineering or procurement tasks) may be split across multiple CWAs, increasing the number of work packages and deliverables to manage. Additionally, the uniform structure may require additional customization when applied to projects with unique organizational or contract-specific requirements.

4.4.2. Schedule Composition and Activity Volume

A quantitative comparison of schedule components across the Traditional, AWP-M1, and AWP-M2 models is provided in

Table 17. Schedule activity quantity comparison summary.

Attribute		Traditional Schedule (A)	AWP-M1 (B)			AWP-M2 (C)
			B	B-A	B/A	C	C-A	C/A
ENG	AE/BOP	24	159	135	6.63	159	135	6.63
	NSSS/RCB	31	55	24	1.77	55	24	1.77
	Common	10	2	−8	0.20	2	−8	0.2
	Interface/IEWP/ENG-IF	98	203	105	2.07	164	66	1.67
	Sum	163	419	256	2.57	380	217	2.33
PRO	AE/BOP	-	87	87	-	87	87	-
	NSSS/RCB	9	33	24	3.67	33	24	3.67
	Common	24	-	−24	-	-	−24	-
	Interface	59	-	−59	-	-	−59	-
	Sum	92	120	28	1.30	120	28	1.30
CON	AE/BOP	89	101	12	1.13	101	12	1.13
	NSSS/RCB	46	46	-	1.00	46	-	1.00
	Common	-	-	-	-	-	-	-
	Interface	124	-	−124	-	-	−124
	Sum	259	147	−112	0.57	147	−112	0.57
Sum		517	686	169	1.33	647	130	1.25

. Notable differences are observed in the number of activities, deliverables, and structural nodes required to support each scheduling approach.

In the engineering domain, AWP schedules demonstrate a significant increase in the number of activities. For AE/BOP engineering, the number of activities in both AWP-M1 and AWP-M2 increased by a factor of 6.63 compared to the Traditional model. This surge is attributed to the following:

• The segmentation of engineering deliverables by CWA;
• The replication of identical design activities across multiple CWPs;
• The introduction of Issuance Engineering Work Package (IEWP) and ENG-IF nodes to coordinate shared deliverables across CWPs.

In the NSSS/RCB domain, the activity count increased by a factor of 1.77, primarily due to the redistribution of previously “common” activities into specific CWAs. This shift enables more localized tracking but increases the total volume of tasks.

AWP introduces 133 (M1) and 94 (M2) IEWP activities, which are not present in the Traditional schedule. These elements enhance schedule control by tracking the readiness of EWP sets instead of relying solely on individual engineering tasks.

The AWP models also eliminate the use of traditional interface activities—98 in engineering, 59 in procurement, and 124 in construction—through the adoption of an integrated schedule. This results in more transparent dependencies but requires a higher number of direct relationships between domain-specific activities.

Procurement activities increased significantly, from 33 activities (9 + 24) in the Traditional schedule to 120 activities in both AWP-M1 and AWP-M2, representing approximately a fourfold increase. This increase reflects the redistribution of materials previously grouped under common scopes into individual CWAs.

In construction, the increase in activity volume was relatively modest (about 13%) in the AE/BOP domain and negligible for NSSS/RCB. This reflects the fact that Traditional construction plans already accounted for zonal execution strategies. The additional activity count in AWP reflects further breakdown by physical area (CWA), such as by level or building.

Overall, the total number of activities increased from 517 in the Traditional schedule to 686 (AWP-M1) and 647 (AWP-M2). The number of engineering deliverables increased from 269 to 940, with 202 and 163 work packages, respectively, for AWP-M1 and M2—compared to only 48 in the Traditional model.

4.4.3. Schedule Logic and Relationship Density

Table 18. Schedule relationship comparison summary.

Predecessor ⇒ Successor	Traditional Schedule (A)	AWP-M1 (B)			AWP-M2 (C)
		B	B-A	B/A	C	C-A	C/A
ENG ⇒ ENG	102	550	448	5.39	550	448	5.39
ENG ⇒ PRO	22	40	18	1.82	40	18	1.82
ENG ⇒ CON	174	153	−21	0.88	127	−47	0.73
PRO ⇒ PRO	22	80	58	3.64	80	58	3.64
PRO ⇒ CON	74	40	−34	0.54	40	−34	0.54
CON ⇒ CON	193	214	21	1.11	214	21	1.11
Total	589	1079	490	1.83	1053	464	1.79

summarizes the schedule logic relationships embedded within the Traditional, AWP-M1, and AWP-M2 schedules. AWP implementation led to a significant increase in logical connectivity, especially in engineering and procurement activities, due to more granular structuring and the introduction of intermediate coordination nodes.

• Engineering-to-Engineering Relationships:

In the Traditional schedule, there were 102 engineering-to-engineering relationships. With AWP, this figure rose to 550—an increase of 5.39 times—due to the following factors:

• Segmentation of engineering deliverables by CWA;
• Introduction of ENG-IF (Engineering Interface) activities to coordinate replicated deliverables across CWPs;
• Addition of IEWP (Issuance of Engineering Work Package) activities that structure the design completion status per EWP.

• Engineering-to-Procurement Relationships:

The Traditional schedule contained 22 engineering-to-procurement relationships, which were implemented via interface activities. These interface activities exist on both the engineering and procurement side of the schedule, so the actual number of logical connections between engineering and procurement activities can be interpreted as approximately 11.

In contrast, AWP explicitly models 40 engineering-to-procurement relationships based on material groupings and CWA alignment. This is nearly four times higher than the actual interdependency count in the Traditional schedule, highlighting the granularity introduced by AWP.

• Engineering-to-Construction Relationships:

Similarly, the Traditional schedule includes 174 connections between engineering and construction. However, about 87 of these represent mirrored interface activities on each side, meaning that only half reflect true logical dependencies.

AWP replaces interface activities with IEWPs nodes, which serve as centralized coordination points. AWP-M1 and M2 shows 96 and 70 IEWP → CON relationships, respectively, which is similar in magnitude to the relationships in the Traditional schedule but offers better modularity and traceability.

• Procurement-to-Procurement Relationships:

Procurement-to-procurement dependencies increased from 22 to 80, a 3.64× rise, due to the breakdown of bulk material activities by CWA.

• Procurement-to-Construction Relationships:

Procurement-to-construction relationships decreased from 74 (Traditional) to 40 (AWP). Again, if half of Traditional connections are interpreted as mirrored interfaces, then the effective connection count is 37—roughly equivalent to AWP’s explicit 40 dependencies.

• Construction-to-Construction Relationships:

Construction-to-construction links increased modestly from 193 to 214 in AWP models (1.11×), mainly due to the further segmentation of previously common or building-wide activities into CWA-specific tasks.

This increase in the number of relationships does not merely indicate added complexity, but rather reflects a more detailed and structured planning approach enabled by AWP. The rise in relationship volume demonstrates improved modeling of dependencies between tasks at the work package level. These enhanced linkages support better workface readiness, enable modularized execution, and establish clearer handover structures. Consequently, they contribute to improved schedule reliability and greater alignment between planning and field execution.

4.4.4. Delay Simulation and Mitigation Effectiveness

To quantitatively assess the schedule-level benefits of applying the AWP framework, construction productivity and workface planning adjustments were introduced into the baseline schedules.

Table 19. Comparison due to construction productivity adjustment.

Schedule	Finish Date	Schedule Variance
Traditional	11 June 2029	ⓐ
AWP applied (25% productivity + 90-day lag)	12 April 2029	ⓐ—60 days

presents the comparison between the traditional schedule and the AWP-applied schedule under the assumption of a 25% improvement in construction productivity and the implementation of a 90-day lead time requirement for engineering deliverables prior to field execution. This adjustment was applied to both AWP-M1 and AWP-M2 schedules, and the resulting project finish dates were compared against the traditional baseline schedule.

This result demonstrates that, even with the imposition of a 90-day buffer requirement before CWP execution, the 25% productivity gain led to a net schedule improvement of 60 days. This highlights AWP’s potential to accelerate overall project completion when workface planning is systematically implemented in conjunction with construction zone alignment and early deliverable readiness.

In scenario ①, a 360-day delay was introduced individually to selected engineering activities across disciplines. These delays were applied directly to the designated activities and not at the IEWP level. In the traditional schedule, the critical path initially ran through RCB-related construction activities and remained unchanged even after the delays—the impacted activities extended the overall duration without shifting the path logic.

However, in both AWP-M1 and AWP-M2 schedules, where the initial critical path also passed through RCB-related construction, the same delays led to a shift in the critical path toward the delayed activities located in CWA2. This shift was observed as shown in

Table 20. 360-day delay and Scenario ① mitigation result comparison summary.

Delay Activity			Variance on 360-Day Delay				B, C 30-Day Backlog
Traditional (A)	AWP-M1 (B)	AWP-M2 (C)	A-Base (ⓐ)	B,C-Base (ⓑ)	ⓑ − ⓐ	Critical Path Changes on ⓑ
310C118PI	CWP01-EWP10-3101C118PI	CWP01-EWP9-3101C118PI	270	270	0	-	−30
310E145PI	CWP03-EWP16-3101E145PI	CWP03-EWP13-3101E145PI	209	222	13	Overall Construction → Cable Tray	−30
310J159PI	CWP06-EWP24-3101J159PI	CWP06-EWP21-3101J159PI	0	0	0	-	0
310P193PI	CWP07-EWP27-3101P186PI	CWP03-EWP25-3101P193PI	270	270	0	-	−30
320C118PI	CWP09-EWP41-3202C118PI	CWP09-EWP31-3202C118PI	173	217	44	RCB → CWA2	−30
320E145PI	CWP00-EWP46-3202E145PI	CWP00-EWP35-3202E145PI	172	261	89	RCB → CWA2 Activity(E3) Duration difference	−30
320J159PI	CWP12-EWP52-3202J159PI	CWP12-EWP40-3202J159PI	0	0	0	-	0
320P193PI	CWP10-EWP57-3202P193PI	CWP10-EWP44-3202P193PI	173	217	44	RCB → CWA2	−30

:

• 320C118PI, 320E145PI, and 320P193PI: The delays caused the critical path to move from RCB-related construction to activities in CWA2.
• In the case of 310E145PI, while the traditional schedule had a direct link from engineering to overall construction, the AWP schedules restructured the sequence such that engineering delay propagated through procurement and construction, particularly through the cable tray activity chain in CWA2.

Other activities like 310J159PI and 320J159PI showed no delay impact and stayed outside the critical path in all scenarios.

When the lead time was reduced—allowing for a 30-day float within the IEWP preparation buffer—all delay-impacted scenarios showed a corresponding 30-day recovery. This demonstrated that the AWP framework’s segmentation and buffering capability can absorb localized delays when managed appropriately, offering resilience against schedule disruptions.

In Scenario ②, the same set of discipline-specific activities used in Scenario ① were restructured and grouped into EWPs, with the corresponding 360-day delays applied at the I-EWP. This scenario aimed to test whether the AWP structure’s logical segmentation and re-sequencing flexibility could mitigate delays more effectively under a centralized delay input.

The results, summarized in

Table 21. Scenario ② mitigation result comparison summary.

Delay Activity		Variance on 360-Day Delay Finish Date		Variance on Post Re-Sequencing Delay Finish Date (Mitigated Dates)
AWP-M1 (A)	AWP-M2 (B)	A-Baseline	B-Baseline	A-Baseline	B-Baseline
I-EWP10	I-EWP9	186	247	186 (0)	247 (0)
I-EWP16	I-EWP13	0	0	0	0
I-EWP24	I-EWP21	0	0	0	0
I-EWP27	I-EWP25	34	185	0	0
I-EWP41	I-EWP31	124	185	101 (−23)	111 (−74)
I-EWP46	I-EWP35	0	0	0	0
I-EWP52	I-EWP40	0	0	0	0
I-EWP57	I-EWP44	170	231	96 (−74)	157 (−74)

, revealed the following patterns:

• RCB-zone delays (I-EWP10, I-EWP27):

  ✓

  These elements, associated with single-zone execution in the RCB, showed no recovery after re-sequencing, as there were no alternative packages or CWAs with the same work types scheduled earlier or in parallel. The lack of flexibility in spatial and temporal overlap restricted mitigation.

• CWA2-zone delays (I-EWP41, I-EWP57 in M1; I-EWP31, I-EWP44 in M2):

  ✓

  These were able to make significant delay mitigation through re-sequencing.

  ✓

  For instance, 23 days to 74 days reduction after re-sequencing.

  ✓

  This was possible due to the availability of follow-up CWAs (e.g., CWA3, CWA4, CWA5) with similar work scopes that allowed flexible scheduling.

• Non-impacted or high-float activities (I-EWP16, 24, 46, 52 for M1; I-EWP13, 21, 35, 40 for M2):

  ✓

  These showed no schedule impact due to adequate float or already being outside the critical path.

Overall, Scenario ② demonstrated that the AWP structure’s segmentation by CWAs and the flexibility of I-EWPs support delay mitigation through workface-level re-sequencing, particularly in areas with overlapping or sequential work packages.

These findings highlight the importance of spatial redundancy and float buffering in AWP planning, especially for disciplines executed across multiple CWAs.

5. Discussion

• The results indicate that AWP-based configurations exhibit higher schedule resilience than traditional CPM logic, particularly through robustness against early engineering delays and rapidity of recovery via IWP re-sequencing [14,15].
• The assumption of a 25% construction-productivity uplift is a simplification. It was adopted as a literature-based scenario value to enable transparent comparison and does not represent trade- or site-specific measurements. Future studies should calibrate differentiated uplifts (e.g., concrete, rebar, MEI) using validated field data or formal meta-analysis.
• Procurement effects were represented deterministically through their intermediate placement between engineering and construction, rather than through stochastic risk distributions. This simplification may understate uncertainty, and probabilistic procurement-risk modeling should be considered in future work.
• The hypothetical model represents 22–28% of referenced plant volume. While the structure was derived from standardized FSAR data to preserve plausibility, no sensitivity runs were performed. As the study’s purpose was comparative, not predictive, results should be interpreted as relative responses. The model was physically simplified and scaled down to serve the needs of a preliminary study, which included treating the RCB as a single construction zone. Accordingly, future research should validate zone-level segmentation against complete, real NPP project schedules where data are accessible.
• The reported 60-day net improvement is a deterministic result derived from fixed scenario logic. Because no probabilistic sampling was conducted, no confidence interval or p-value can be provided. This value should therefore be interpreted as illustrative of structural schedule effects, not as a statistically validated result. Future research should apply stochastic risk analysis to test the statistical robustness of such improvements.
• Commissioning and regulatory gates were not included due to scope and confidentiality. However, research is emerging on how AWP may be applied in commissioning, particularly on reconciling physical packages used in construction with system-level packages required for system turnover. This represents a future extension of the current study.
• While this study concentrated on schedule-level impacts, the adoption of AWP also requires procedural, administrative, and system changes across design, procurement, and project management processes. These adjustments inevitably involve additional time and cost, which were not analyzed here. ROI should therefore not be inferred from schedule benefits alone. Future research should extend the analysis to include these broader dimensions, so that cost–benefit trade-offs are evaluated in a comprehensive manner.
• The weighting applied in

  Table A2. Advanced work packaging assessment matrix.

  No.	Criteria	Method 1		Method 2		Method 3		Method 4
  		Score	Rationale	Score	Rationale	Score	Rationale	Score	Rationale
  1	Scope Definition of EWP	5	Discipline/function-based packaging; sub-discipline and function specific range of work.	5	Discipline-based packaging; discipline specific range of work.	3	Discipline/phase-based packaging; discipline specific range of work; may require overlapping phases of work	5	Discipline-based packaging; discipline specific range of work.
  2	Scope Definition of PWP	5	Type-based packaging	5	Type-based packaging	3	Discipline (Multi-material type)/phase-based packaging; compatibility with PO & process needed.	4	Discipline (Multi-material type)-based packaging; compatibility with PO needed.
  3	WBS Compatibility of EWP	5	Aligns with WBS hierarchy.	5	Aligns with WBS hierarchy.	4	Can align with WBS hierarchy; but duplicated functions required	5	Aligns with WBS hierarchy.
  4	WBS Compatibility of PWP	5	Aligns with WBS hierarchy.	5	Aligns with WBS hierarchy.	4	Can align with WBS hierarchy; but duplicated types required	5	Aligns with WBS hierarchy.
  5	EWP as a Schedule Unit	5	Compliant as an independent schedule unit; may require supplementary activity relationships.	4	Compliant as an independent schedule unit; but require supplementary activity relationships.	3	Compliant as an independent schedule unit; but require supplementary activity relationships and overlapping for phased functions.	4	Compliant as an independent schedule unit; but require supplementary activity relationships.
  6	PWP as a Schedule Unit	4	Compliant as an independent schedule unit; may require supplementary activity/relationship; require a separate plan for field-requested dates	4	Compliant as an independent schedule unit; may require supplementary activity/relationship; require a separate plan for field-requested dates	3	Compliant as an independent schedule unit; require supplementary activity/relationship; require a separate plan for field-requested dates; overlap in phased procurement; require compatibility with PO.	3	Compliant as an independent schedule unit; require supplementary activity/relationship; require a separate plan for field-requested dates; require compatibility with PO.
  7	EWP Compatibility with Schedule Hierarchy	4	Appropriate and align with WBS hierarchy; but may generate an extensive number of work packages	5	Appropriate and align with WBS hierarchy.	3	Appropriate and align with WBS hierarchy; but may generate an extensive number of work packages as the phases increase	5	Appropriate and align with WBS hierarchy.
  8	PWP Compatibility with Schedule Hierarchy	4	Appropriate and align with WBS hierarchy; but may have extensive number of work packages	4	Appropriate and align with WBS hierarchy; but may have extensive number of work packages	3	Appropriate and align with WBS hierarchy; but may generate an extensive number of work packages as the phases increase; require compatibility with PO.	4	Appropriate and align with WBS hierarchy; require compatibility with PO.
  9	Constructability of EWP	5	Align with and support the POC	5	Align with and support the POC	5	Align with and support the POC	5	Align with and support the POC
  10	Constructability of PWP	4	Align with and support the POC; however, a separate material delivery management plan is needed to prevent a large amount of inventory from being stored on-site for a long period.	4	Align with and support the POC; however, a separate material delivery management plan is needed to prevent a large amount of inventory from being stored on-site for a long period.	3	Align with and support the POC; separate material delivery management plan is needed to prevent a large amount of inventory from being stored on-site for a long period; require compatibility with PO.	3	Align with and support the POC; separate material delivery management plan is needed to prevent a large amount of inventory from being stored on-site for a long period; require compatibility with PO.
  11	Progress Measurement by EWP	5	Measurable.	4	Measurable; but require separate plan to track specific tasks	4	Measurable; but require separate plan to track specific tasks	4	Measurable; but require separate plan to track specific tasks
  12	Progress Measurement by PWP	5	Measurable.	5	Measurable.	3	Measurable; but require separate plan to track specific tasks; require compatibility with PO.	3	Measurable; but require separate plan to track specific tasks; require compatibility with PO.
  Score total /Converted Score		56/93.6%		55/92%		41/69.2%		50/83.2%

  was developed by the first and corresponding author, both with 10–20 years of project management experience, and then informally cross-checked with a small group of senior practitioners. Because direct AWP implementation experience in nuclear projects is not yet available, no inter-rater reliability testing was possible. This is recognized as a limitation, and future work should validate such weighting through structured expert panels and reliability methods.
• This study applied deterministic, step-wise delays in order to compare schedule-structure responses under controlled conditions. The use of standardized nuclear project schedule data made deterministic testing more suitable than probabilistic sampling. While this approach was appropriate for a preliminary comparison, it does not provide delay distributions or confidence intervals. Future studies should apply stochastic methods such as Monte Carlo or PERT to complement the findings and validate statistical robustness.
• NPP schedules are highly sensitive to external uncertainties, including policy changes, regulatory approval processes, and supply chain disruptions. These factors were not quantified in this study, as the analysis focused on deterministic, comparative scenarios to illustrate AWP-specific effects. Nonetheless, such uncertainties exert significant influence on actual project performance. Future studies should extend the analysis to incorporate probabilistic schedule risk assessments (e.g., Monte Carlo approaches), so that the impact of policy, regulatory, and supply chain risks can be evaluated in a more comprehensive manner.
• As authors’ best knowledge, no published nuclear power project has reported full-scale application of AWP. International references remain at the level of conceptual proposals or emerging contractual requirements rather than validated case studies. In several recent projects, owners have signaled an interest in incorporating AWP principles into execution planning. Against this backdrop, the present work is positioned as a preliminary study to provide a first step toward evaluating the potential of AWP in nuclear construction.

6. Conclusions, Limitations, and Recommendations

6.1. Conclusions

This study investigated the application of Advanced Work Packaging (AWP) to nuclear power plant (NPP) construction scheduling, contrasting it with the Korean traditional practices. Key conclusions are as follows:

• In evaluating four packaging schemes against six criteria (scope definition, WBS compatibility, schedule-unit suitability, fit with schedule hierarchy, constructability, and progress measurement), Methods 1 and 2 consistently scored 4–5 and were advanced. Method 1 uniquely earned top marks for “EWP as a Schedule Unit” and “Progress Measurement by EWP,” because FBS-based EWPs are self-contained and auditable and pair cleanly with equipment-type PWPs. Method 2 uniquely earned top marks for “EWP Compatibility with Schedule Hierarchy,” as discipline-based EWPs map directly to the OBS within the WBS. By contrast, Methods 3–4 were not advanced: both exhibit PWP configurations that conflict with PO lifecycles (composition and process), demand supplementary activity relationships and field-requested-date plans, and require management plans at equipment-level traceability; in addition, Method 3’s procurement-phase split introduces temporal overlap that further fragments schedule units and progress measurement.
• WBS Structure and Work Packaging: Traditional WBS structures varied by EPC phase (contract-based, discipline-based, package-based, etc.), creating challenges for integrated management and reporting. By contrast, the AWP-based WBS consistently followed: Project → CWA → Phase (E/P/C) → Discipline (OBS) → WP. This structure expanded work packages from 48 (traditional) to 202 (AWP-M1) and 163 (AWP-M2), with engineering deliverables standardized at 940. The increase is explained by CWA-level scoping that replicates discipline deliverables by area; this improves area-based traceability and reporting but raises management complexity due to package proliferation.
• Schedule Activities and Relationships: AWP schedules significantly expanded activities compared to the traditional baseline.

  ✓

  Engineering (AE/BOP) increased 6.6×, and NSSS/RCB about 1.8×, due to CWA segmentation and replicated activities.

  ✓

  Procurement activities increased almost 4× (33 → 120) as previously common items were allocated to CWAs.

  ✓

  Construction activities showed minor increases since traditional schedules already applied area-based planning.

  ✓

  Relationships expanded substantially: ENG–ENG links increased 5.4×, and ENG–PRO nearly 2×. Traditional interface activities (281) disappeared, replaced by IEWP and ENG-IF logic, providing stronger integration across phases.

• Schedule Impacts: Applying a 25% construction productivity improvement shortened durations by about 120 days, while the 90-day lag for EWP issuance included 60 days of readiness for IWP development, which offset part of the gains. The net outcome was a 60-day earlier project completion compared to the traditional baseline.
• Delay Simulation and Mitigation: In Scenario ① (360-day engineering delays), the traditional critical path remained in RCB-related construction, while AWP shifted the critical path toward delayed activities in CWA2, demonstrating segmentation-driven responsiveness. In Scenario ② (re-sequencing), substantial recovery was observed in successive CWAs (e.g., I-EWP41 mitigated 23 days, I-EWP57 recovered 74 days). These results confirm that AWP’s zonal structure enables localized delay absorption when alternative workfaces exist. Recovery concentrated in the AB because spatial redundancy across adjacent CWAs enabled parallel advancement of similar work types, whereas the RCB’s contiguous, critical structural scope offered few viable substitutes.
• These conclusions should be interpreted with caution, as they are derived from a simplified, scaled model of the Nuclear Island, applied under generalized productivity and lead-time assumptions, with delays treated deterministically. While sufficient for a preliminary study, these boundaries limit direct generalization to full-scale NPP projects.

6.2. Limitations

• Simplified Modeling Assumptions: The schedules developed in this study relied on a simplified standard work logic and assumptions rather than fully reflecting the detailed complexities of an actual project. For example, construction productivity was adjusted uniformly by 25%, and 90-day engineering/procurement lead time were modeled in a generalized manner. Such simplifications limit the direct applicability of the results to real projects, where variability and uncertainty are more pronounced. Furthermore, the AWP-based WBS was structured around CWAs, EWPs, PWPs, and CWPs in accordance with industry guidelines. However, not all possible variations of work package configurations were considered. For instance, procurement strategies were simplified by excluding detailed vendor-specific practices and fabrication steps that could significantly influence the applicability.
• Limited Risk Coverage: The delay simulations were designed around selected engineering activities across disciplines. While effective for comparative analysis, they did not incorporate the broader range of risks that typically affect nuclear projects, such as regulatory reviews, vendor-related delays, or site-specific conditions. The mitigation effects observed should therefore be interpreted within this limited scope.
• Re-Sequencing Assumptions: In this study, Scenario ② re-sequencing was applied only to AWP schedules, under the assumption that such adjustments are uniquely aligned with WFP practices. However, in reality, similar resequencing and adjustments are also carried out in traditional project execution once construction readiness allows. This limitation may have led to an overestimation of the differential benefit attributed solely to AWP.

6.3. Recommendations for Future Research

• Detailed Schedule Development: Current industry practice using Level 3 IPS (Integrated Project Schedule) or CPS (Critical Path Schedule) provides activity relationships but lacks the level of granularity required for accurate critical path analysis. Future research should develop model-based detailed schedules in which CWAs, POCs and IWPs are explicitly linked to 4D/5D BIM objects, enabling higher-precision critical path monitoring, quantity-driven progress validation, and automated look-ahead generation.
• Stakeholder Engagement at Project Initiation: Successful AWP implementation requires active involvement of procurement and construction stakeholders from the project outset. Future research should investigate contractual and procedural approaches that move away from the traditional reliance on fixed interface activities, enabling earlier alignment of deliverables and responsibilities.
• FOAK (For first of a kind) Considerations: Adopting AWP in FOAK settings requires rethinking how engineering deliverables, procurement packages, and construction packages are defined. Packages should be BIM-addressable from the start of the project (e.g., derive the Path of Construction from the model; tie fabrication lots to model assemblies). Future work should also assess the practical and business impacts on EPC delivery—such as data governance, change control, and traceability.

6.4. Contributions of the Study

This research contributes to both academia and industry in the following ways:

• Applying the AWP Project Definition Assessment Tool to a Korean nuclear EPC context, identifying current alignment and gaps.
• Demonstrating a practical methodology for constructing AWP-based schedules and comparing them against traditional schedules using Primavera P6.
• Quantitatively analyzing differences in work packages, deliverables, activities, and relationships, clarifying the impacts of AWP adoption.
• Testing delay simulation scenarios, showing how AWP segmentation and re-sequencing provide resilience to schedule disruptions.
• Offering practical recommendations on WBS standardization, tool integration, risk-driven planning, and FOAK implications, bridging theoretical research with industrial application.

Ultimately, this study shows that adopting AWP in NPP construction schedules can enhance consistency, traceability, and resilience, while also highlighting the challenges and research directions needed for successful implementation.