A Simulation-Based Optimisation for Contractors in Precast Concrete Projects
10th Nordic Conference on Construction Economics and Organization
eISBN: 978-1-83867-051-1
ISSN: 2516-2853
Publication date: 1 May 2019
Abstract
Purpose
This paper aims to provide decision support for precast concrete contractors about both precast concrete supply chain strategies and construction configurations.
Design/Methodology/Approach
This paper proposes a simulation-based optimisation for supply chain and construction (SOSC) during the planning phase of PC building projects. The discrete event simulation is used to capture the characteristics of supply chain and construction processes, and calculate construction objectives under different plans. Particle swarm optimisation is combined with simulation to find optimal supply chain strategies and construction configurations.
Findings
The efficiency of SOSC is compared with the parametric simulation approach. Over 70 per cent of time and effort used to simulate and compare alternative plans is saved owing to SOSC.
Research Limitations/Implications
Building simulation model costs a lot of time and effort. The data requirement of the proposed method is high.
Practical Implications
The proposed SOSC approach can provide decision support for PC contractors by optimising supply chain strategies and construction configurations.
Originality/Value
This paper has two contributions: one is in providing a decision support tool SOSC to optimise both supply chain strategies and construction configurations, while the other is in building a prototype of SOSC and testing it in a case study.
Keywords
Citation
Chen, S., Feng, K. and Lu, W. (2019), "A Simulation-Based Optimisation for Contractors in Precast Concrete Projects", Lill, I. and Witt, E. (Ed.) 10th Nordic Conference on Construction Economics and Organization (Emerald Reach Proceedings Series, Vol. 2), Emerald Publishing Limited, Leeds, pp. 137-145. https://doi.org/10.1108/S2516-285320190000002019
Publisher
:Emerald Publishing Limited
Copyright © 2019, Shiwei Chen, Kailun Feng, Weizhuo Lu.
License
Published in the Emerald Reach Proceedings Series. Published by Emerald Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial and non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode
1. Introduction
Precast concrete (PC) construction is a construction method in which building components are produced in an off-site factory and transported to construction site to be assembled into a building. In conventional cast-in-situ concrete projects, contractors decide construction configurations, including the types and quantities of equipment and workers used for every construction activity, to make construction plans. Because many parts of a PC building are produced in factories rather than construction site, PC supply chain has great impact on the construction objectives (Zhai et al., 2013). Thus, when making construction plans of PC projects, contractors should consider not only construction configurations, but also PC supply chain strategies, including whether and where to store PC components.
In PC projects, contractors usually decide PC supply chain strategies and construction configurations according to a rule of thumb, leading to extra costs, prolonged construction duration or worse sustainable performance (Pheng and Chuan, 2001). Previous studies analysing PC supply chain strategies are mostly conducted from the perspective of PC component suppliers and seldom combine construction and PC supply chain together (Hosseini et al., 2018). There still lacks a decision-support method selecting both PC supply chain strategies and construction configurations for contractors.
When considering PC supply chain with construction, it is difficult to compare all alternative plans manually because the interactions between PC supply chain and construction are dynamic and complex (Arashpour et al., 2017). Progress of construction activities decides when PC components are needed on construction site and PC component supply chain influence PC-related construction processes. In addition, there are too many combinations of PC supply chain strategies and construction configurations. Finding the optimal solutions manually takes too much time and effort, which is unacceptable in the actual construction practice (Nguyen et al., 2014).
Computer-based simulation has the ability of capturing the dynamic and complex interactions among construction engineering systems (AbouRizk, 2010). However, using conventional parametric simulation approach to compare all alternative plans still costs a lot of time because it has to run all possible combinations (Nguyen et al., 2014). Thus, in some studies, simulation is combined with optimisation as an integrated method, known as the simulation-based optimisation (SO), to accelerate the process of finding optimal solutions. The SO method is mainly used in building performance comparison of different designs but seldom used in PC construction area, especially not in problems considering both PC supply chain and construction.
This paper combines discrete-event simulation (DES) and particle-swarm optimisation (PSO) to propose a simulation-based optimisation of supply chain and construction (SOSC) for contractors in PC projects. The SOSC approach can provide decision support for contractors by quickly selecting optimal combinations of PC supply chain strategies and construction configurations from a great number of alternative plans. A prototype of SOSC is developed, and a case study is conducted to demonstrate the effectiveness and efficiency of the proposed method.
2. Precast concrete supply chain strategies
To choose PC supply chain strategy, the contractors should decide whether and where to store PC components after components left factories (Pheng, C. and Chuan, L. 2001). Based on the decisions, PC supply chain strategies can be categorised into three types: just-in-time (JIT) strategy (no storage), on-site storage strategy and off-site storage strategy. An interview with 14 professionals from 3 contractors and 2 suppliers in China is conducted to find out the processes of the three strategies (shown in Figure 1).
In the JIT strategy, PC components are directly transported to construction site and will be unloaded, checked and hoisted as soon as they reach site. The contractor will call PC suppliers for components according to construction progress. The quality of components will be checked when they reach the site, and the components with quality flaws will be returned to the factory for repair.
In on-site storage, the contractor will use components in storage when the components are needed. The contractor will set a recalling point for the number of components and call PC supplier for components when the number of stored components is below recalling point. Owing to limited space on site, the storage usually can only store limited components.
The off-site storage strategy has the same processes as the on-site storage strategy, except the storage is located beyond construction site. Thus, there is usually no limitation for store capacity. The transportation from off-site storage to construction site is usually conducted by the contractor, not PC supplier.
3. SOSC approach
This paper proposes a simulation-based optimisation of supply chain and construction (SOSC) during the planning phase of PC building projects, to provide decision support for contractors. The framework of SOSC is shown in Figure 2.
First, some detailed data about the construction project is collected and analysed, including interactions between construction and supply chain activities, construction sequences, and PC components, equipment and labour needed for each activity.
Based on the collected data, a DES model is built. An Activity-Component-Resource-Action-Sequence (CARS) model proposed by Fischer et al. (1999) is used to capture the characteristics of PC supply chain and construction processes: “Activity” is the PC supply chain and construction activities; “Component” refers to the buildings parts, including PC components and cast-in-situ concrete; “Resource” means construction equipment and workers; “Action” means executed work information for each activity; and “Sequence” is the logic restriction between activities of supply chain and construction. The project data is categorised and input into DES according to CARS.
After building the DES model, two tests need to be conducted. One is minimum simulation runs determination (Lee et al., 2015), which tests reliability of the simulation outcomes under different simulation runs to find out the minimum simulation runs needed to obtain reliable simulation outcomes. Another test is simulation results validation (Lee et al., 2015), which tests the accuracy of DES. The actual PC supply chain strategies and construction configurations are input into the built DES model. The simulation outcomes are compared with actual construction objectives to test the model accuracy. If the accuracy is acceptable, the DES model can be used in next procedure. Otherwise, the model needs modification until it can pass the tests.
Finally, a DES-based PSO (see Figure 3) is used to optimise supply chain strategies and construction configurations. The searching scope is defined to input the alternative plans to PSO, and PSO will generate particles at initial positions to represent initial random plans. Then, these plans are input into DES model to get the simulated construction objectives of each plan, and the objectives will be input into PSO as the fitness value of particles. PSO will select particles with best fitness value and check the convergence criteria. If the convergence criteria or the maximum number of iterations has been reached, the optimisation will stop and combinations of supply chain strategies and construction configurations represented by the selected particles will be output as Pareto solutions. Otherwise, PSO will generate new iteration of combinations and particles will move to new positions accordingly. The particle movements in PSO are based on their own best position in the search-space and the best position found by the entire swarm. The new found improved positions will replace the former local best positions. After particle movement, the PSO will restart from calculating fitness value by simulation and repeat the whole process until the convergence criteria or the maximum number of iterations are reached.
4. Case study
4.1. Case background
A prototype of SOSC is built to demonstrate its effectiveness and efficiency. A precast concrete project in Shenzhen, China, is chosen as the study case because it has the mostly used PC structure in China, the PC shear wall structure, and a high prefabrication rate of 49.5 per cent. Data of the project is collected through reading construction records and planning documents, field survey and interview with the contractor and the supplier.
4.2. Building simulation model
To simulate the PC supply chain and construction system of the studied case, a DES model is built on the platform SIMIO. According to CARS, the construction sequences, interactions between PC supply chain and construction, PC supply chain strategies, construction configurations and performance calculation equations are input into SIMIO to build the DES model.
4.2.1. Inputting project data into DES
According to our field survey and interview with the contractor, the construction sequences and interactions between construction and PC supply chain are summarised and built in SIMIO (see Figure 4). The actual used and alternative PC supply chain strategies and construction configurations (see Table 1) are collected from our interview and document reading. These plans are set as input parameters in the built DES model. In addition, some transportation data, such as the transportation distance, speed, quality failure ratio and average delay time, are also collected through our interview and input into DES.
Alternative and actual used construction plans.
| Tasks | Actual used plans | Alternative plans | Remarks | |
|---|---|---|---|---|
| Components loading & unloading | 3 forklifts (CPC(Q)(Y)D50) | 1∼3 forklifts (CPC(Q)(Y)D50) | 5T | |
| 1∼3 forklifts (CPC(Q)(Y)D60) | 6T | |||
| 1∼3 forklifts (CPC(Q)(Y)D70) | 7T | |||
| PC components hoisting & installation | 70 PC workers | 65, 70, 75 or 80 PC workers | ||
| 2 cranes (STT293) | 2 cranes (STT293) | |||
| 2 cranes (XCP330HG7525) | ||||
| 1 crane (XGT8039) | ||||
| 1 crane (XGT500A8040) | ||||
| Concrete pouring | 10 concrete workers | |||
| 2 concrete pump (HBT6006A-5) | 1∼3 concrete pump (HBT6006A) | 75 kW, 70 m3/h | ||
| 1∼3 concrete pump (HBT8016C) | 132kW, 85m3/h | |||
| 1∼3 concrete pump (HBT6013C) | 90kW, 65m3/h | |||
| Material & labour transportation | 3 construction elevators (SC200/200) | 66 kw, 2 × 2 t | ||
| Rebar processing | 40 rebar processing workers | 20, 30 or 40 rebar processing workers | ||
| 1 rebar bending machine (GK50P) | 4kW | |||
| 1 rebar cutting machine (GJ7-40) | 3kW | |||
| Rebar installation | 20 rebar installation workers | 20, 30 or 40 rebar installation workers | ||
| Formwork installation | 30 formwork installation workers | |||
| Joint grout | 15 joint grout workers | |||
4.2.2. Performance calculation
To compare different plans, three comparison indicators are selected, including construction duration, construction cost and greenhouse gas (GHG) emissions. Construction duration and cost are conventional key objectives, and GHG emissions are important when measuring construction sustainability (Mao et al., 2013). In SIMIO, the construction duration is automatically calculated on the basis of the quantity of work and working productivity. The construction cost in this study includes the buying price and transportation fees of PC components, cost of cast-in-situ materials, cost of construction crews, storage related fees and rental of construction machines. The cost data is found in construction documents. The GHG emissions in this study denotes the GHG emissions of electricity and diesel consumed by construction site operation, PC transportation trucks and construction machines, including rebar processing machines, concrete pumps, cranes, forklifts and construction lifts. The GHG emissions are calculated according to Equation (1).
where G means the total GHG emissions in construction and transportation activities; E means the electricity consumed during construction; D means the diesel consumed during PC component transportation and construction; ge means the GHG emissions factors of electricity, which is 0.714 kg CO2-e/kWh for this case (Provincial Greenhouse Gas Inventory Guidelines, 2011); gd means the GHG emissions factors of diesel, which is 3.153 kg CO2-e/kg (Mao et al., 2013).
4.3. Simulation validation
To examine the reliability and accuracy of the built DES model, the minimum simulation runs determination and simulation outcome validation are conducted. According to our test, the simulated outcomes of all three construction objectives become stable after simulation runs reach 59. Therefore, every simulation in this study will be replicated for 60 times. Then, the actual supply chain strategy (JIT) and construction configurations (see Table 1) are input to the DES model, and simulation outcomes are compared with actual construction objectives. According to our test, the difference between actual construction objectives and average simulation results are all less than 2 per cent, and all actual construction objectives are within the scope of simulation results (between maximum and minimum). This indicates that the built model has acceptable accuracy.
4.4. DES-based PSO
A multi-objective PSO is used on MATLAB to find the Pareto solutions from the alternative plans, and the detailed procedures of calling SIMIO in MATLAB is built based on the SIMIO-MATLAB framework proposed by Dehghanimohammadabadi and Keyser (2017). Table 2 shows some proper parameters for PSO dealing with construction configuration activities used in Wang et al. (2017). The convergence criteria used in this study is the certain number of iterations (10 in this study) with improvement below threshold.
Parameters set in PSO.
| Parameter | Value |
|---|---|
| Population number of each generation (N) | 30 |
| Maximum number of iterations | 500 |
| Acceleration constant (c1) | 0.8 |
| Acceleration constant (c2) | 0.8 |
| Inertia weight (w) | [0.1, 1.2] |
5. Results and discussion
After SOSC, 15 plans are found as Pareto solutions (see Table 3). The construction objectives of Pareto solutions and actual construction plan are shown in Figure 5. According to Figure 5, the construction duration of all optimised plans is less than the actual construction duration. As for GHG emissions or construction cost, not all optimised plans are better than the actual plan. Plans 1, 4, 7, 8, 9, 12 and 15 have less construction duration, less cost and less GHG emissions than the actual plan, which means choosing these plans can get overall improvement. Contractors can choose one plan out of the 15 optimised plans according to their preference in different construction objectives.
Compared to conventional parametric simulation approach, which tests all alternative plans to find out the optimal plans, SOSC is a more efficient method, which saves a lot of time and effort in simulating and comparing alternative plans by reducing the alternative plans needed for simulation. In this case, the search space of all alternative plans is (3 * 3 * 4 * 4 * 3 * 3 * 3 * 3) * 3 = 11,664 combinations of construction configurations * 3 PC supply chain strategies = 34,992 (see Table 1), and only 10,260 combinations have been generated in PSO to find out the optimal plans. According to the reduced simulation runs, (34,992 * 60 − 10,260 * 60)/(34,992 * 60) ≈ 70.68% of time and effort used to simulating and comparing alternative plans has been saved due to SOSC.
Optimal plans.
| Plan no. | Supply chain strategies | Construction configurations | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Forklifts | PC workers | Cranes | Concrete pumps | Rebar processing workers | Rebar installation workers | |||||
| Number | Type | Number | Type | |||||||
| 1 | JIT | 3 forklifts (CPC(Q)(Y)D50) | 65 | 2 | S | 1 | C | 40 | 30 | |
| 2 | 65 | 2 | X | 1 | A | 40 | 40 | |||
| 3 | 65 | 2 | X | 1 | C | 30 | 30 | |||
| 4 | 75 | 2 | S | 1 | A | 20 | 30 | |||
| 5 | 65 | 2 | S | 1 | C | 20 | 40 | |||
| 6 | 75 | 2 | X | 1 | A | 40 | 30 | |||
| 7 | On-site storage | 65 | 2 | S | 1 | A | 20 | 40 | ||
| 8 | 65 | 2 | S | 2 | A | 20 | 30 | |||
| 9 | 65 | 2 | X | 1 | A | 40 | 30 | |||
| 10 | 75 | 2 | X | 3 | C | 30 | 40 | |||
| 11 | 70 | 2 | X | 1 | C | 20 | 30 | |||
| 12 | 70 | 2 | S | 1 | A | 20 | 20 | |||
| 13 | 75 | 2 | S | 3 | A | 30 | 40 | |||
| 14 | 80 | 2 | S | 3 | A | 30 | 40 | |||
| 15 | 65 | 2 | S | 1 | A | 20 | 20 | |||
6. Conclusion
This paper combines DES and PSO and proposes a simulation-based optimisation of PC supply chain and construction (SOSC) for contractors in PC projects. DES is used to capture the dynamic and complex interactions between PC supply chain and construction, and calculate the construction objectives of alternative plans. PSO is used to search for the optimal combinations of PC supply chain strategies and construction configurations for contractors based on the simulated construction objectives. A prototype of SOSC is built and tested in a case study to demonstrate its effectiveness and efficiency. The efficiency of SOSC are compared with the parametric simulation approach, and over 70 per cent of time and effort used to simulating and comparing alternative plans has been saved due to SOSC.
This paper has two contributions: one is in providing a decision support tool SOSC to optimise both supply chain strategies and construction configurations, while the other one is in building a prototype of SOSC and testing it in a case study. The proposed approach also has some limitations. Firstly, building simulation model costs a lot of time and effort. In addition, SOSC has high data requirement. How to overcome these limitations will be studied in the future research.
References
AbouRizk, 2010AbouRizk, S. (2010), “Role of simulation in construction engineering and management”, Journal of Construction Engineering and Management, Vol. 136, No. 10, pp. 1,140–1,153.
Arashpour, Abbasi, Arashpour, Hosseini, and Yang, 2017Arashpour, M., Abbasi, B., Arashpour, M., Hosseini, M. and Yang, R. (2017), “Integrated management of on-site, coordination and off-site uncertainty: Theorizing risk analysis within a hybrid project setting”, International Journal of Project Management, Vol. 35, pp. 647–655.
Dehghanimohammadabadi, and Keyser, 2017Dehghanimohammadabadi, M. and Keyser, T. (2017), “Intelligent simulation: Integration of SIMIO and MATLAB to deploy decision support systems to simulation environment”, Simulation Modelling Practice and Theory, Vol. 71, pp. 45–60.
Fischer, Aalami, Kuhne, and Ripberger, 1999Fischer, M., Aalami, F., Kuhne, C. and Ripberger, A. (1999), “Cost-loaded Production Model for Planning and Control”, Durability of Building Materials and Components, Vol. 8, No. 4, pp. 2,813–2,824.
Lee, Filatova, Ligmann-Zielinska, HassaniMahmooei, Stonedahl, Lorscheid, Voinov, Polhill, Sun, and Parker, 2015Lee, J., Filatova, T., Ligmann-Zielinska, A., HassaniMahmooei, B., Stonedahl, F., Lorscheid, I., Voinov, A., Polhill, G., Sun, Z. and Parker, D., (2015), “The complexities of agent-based modeling output analysis”, Journal of Artificial Societies and Social Simulation, Vol. 18, No. 4, pp. 1–26.
Mao, Shen, Shen, and Tang, 2013Mao, C., Shen, Q., Shen, L. and Tang, L. (2013), “Comparative study of greenhouse gas emissions between off-site prefabrication and conventional construction methods: two case studies of residential projects”, Energy and Buildings, Vol. 66, pp. 165–176.
National Development and Reform Commission (NDRC), 2011National Development and Reform Commission (NDRC). (2011), Provincial Greenhouse Gas Inventory Guidelines, NDRC, Beijing.
Nguyen, Reiter, and Rigo, 2014Nguyen, A., Reiter, S. and Rigo, P. (2014), “A review on simulation-based optimization methods applied to building performance analysis”, Applied Energy, Vol. 113, No. 6, pp. 1,043–1,058.
Pheng, and Chuan, 2001Pheng, C. and Chuan, L. (2001), “A study of the readiness of precasters for just-in-time construction”, Work Study, Vol. 50, No. 4, pp. 131–140.
Wang, Feng, and Lu, 2017Wang, Y., Feng, K. and Lu, W. (2017), “An environmental assessment and optimization method for contractors”, Journal of Cleaner Production, Vol. 142, No. 4, pp. 1,877–1,891.
Zhai, Reed, and Mills, 2013Zhai, X., Reed, R. and Mills, A. (2013), “Factors impeding the off-site production of housing construction in China: an investigation of current practice”, Construction Management and Economics, Vol. 32, pp. 40–52.
This research was supported by Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (Formas), the Swedish Foundation for International Cooperation in Research and Higher Education (STINT).
- Prelims
- THE ECONOMICS AND BUSINESS OF CONSTRUCTION
- Updating and Cleaning Out: The “Make or Buy” Decision in Construction Revisited
- Bispevika Project: Research for Constructing a Collaborative Value Chain
- Social Considerations in the Procurement of Road and Railroad Projects in Sweden
- Standardization and Industrialized Construction of Special Purpose Building
- Identifying Contradictions of Integrating Life-Cycle Costing in Design Practices
- Advancing Networking-Based Business Management in Construction Markets
- Contracts and Culture in a Partnering Project
- Sub-Contractors’ Perception of Contracting: The Case of Crime
- Project Managers: Gatekeepers or Inside Men?
- The Hybridity of Strategic Partnerships and Construction Supply Chain Management
- Dynamic Capabilities and Risk Management: Evaluating the CDRM Model for Clients
- An Opposite Design-Build Procurement Method: Competing on Quality with a Fixed Price
- CONSTRUCTION AND PROJECT MANAGEMENT
- An Appraisal of Water Infrastructure Projects’ Financing Challenges in South Africa
- The Soft Factors in Design Management: a Hidden Success Factor?
- Room to Manoeuvre: Governing the Project Provisions
- A Longitudinal View of Adopting Project Alliancing: Case Finland
- A Simulation-Based Optimization for Contractors in Precast Concrete Projects
- Governed by Municipal Land Allocations: Implications for Housing Developers
- Situation Picture Through Construction Information Management
- Who Benefit from Crime in Construction? A Structural Analysis
- Quality Evaluation of Contractor’s Schedule in the Bidding Phase
- Activity Cruciality as Measure of Network Schedule Structure Resilience
- Construction Programmes and Programming: A Critical Review
- Procurement Research: Current State and Future Challenges in the Nordic Countries
- Exploitative Learning in Inter-Organizational Projects: Evidence from Dutch Infrastructure Practices
- The Transition from Design-Bid-Build Contracts to Design-Build
- Exploring the Dynamics of Supplier Innovation Diffusion
- Understanding Collaborative Working in a Facilitated Interdisciplinary Environment
- Ensuring Successful Knowledge Transfer in Building Renovation Projects
- Public Private Collaboration in the Context of Zero Emission Neighbourhood
- Strategizing and Project Management in Construction Projects: An Exploratory Literature Review
- BUILDING INFORMATION, DATA AND DIGITALIZATION
- BIM-Enabled Education: a Systematic Literature Review
- A BIM-Enabled Learning Environment: a Conceptual Framework
- “I Work All Day with Automation in Construction: I am a Sociomaterial-Designer”
- Developing Smart Services to Smart Campus
- An Overview of BIM Adoption in the Construction Industry: Benefits and Barriers
- BIM for Construction Education: Initial Findings from a Literature Review
- Model for Smart, Self-learning and Adaptive Resilience Building
- Investigating the Drop-Out rate from a BIM Course
- INNOVATIONS IN THE CONSTRUCTION PROCESS
- Senior Residence Concepts in Norway: Challenges and Actions for a Sustainable Development
- 3D-Printing Technology in Construction: Results from a Survey
- Product and Manufacturing Systems Alignment: a Case Study in the Timber House Building Industry
- Opening the Black Box of Accessibility Regulation
- Orchestrating Multi-Actor Collaborative Innovation Across Organizational Boundaries
- SUSTAINABILITY AND RESOURCE EFFICIENCY
- Social Sustainability in Modelling of Value Creation in Housing Refurbishment
- Reviewing the Role of Sustainability Professionals in Construction
- Exploring the Evolution and Impact of Building Environment Assessment Methods in Achieving Green Building
- STAKEHOLDERS OF CONSTRUCTION AND REAL ESTATE
- Challenging the Rhetoric of Construction Briefing: Insights from a Formula 1 Sports Venue
- Underlying Causes for Risk Taking Behaviour Among Construction Workers
- Towards Developing a Framework for User-Driven Innovation in Refurbishment
- Reconstructing Knowledge Integration in the Norwegian AEC-Industry
- Institutional Complexity for Chinese International Contractors
- BUILT ENVIRONMENTS
- BIM Related Innovation in Healthcare Precinct Design and Facilities Management
- Location is Crucial in Retrofit: Strategy Selection in Different Regions
- CONSTRUCTION EDUCATION AND RESEARCH
- From Theoretical to Practical Competence on Health and Safety
- A Test Platform of Viable Methods to Improve Production and Learning on Construction Sites