Multi-Criteria Analysis of a Developed Prefabricated Footing System on Reactive Soil Foundation

Abstract

The need for advancements in residential construction and the hazard induced by the shrink–swell reactive soil movement prompted the development of the prefabricated footing system of this study, which was assessed and compared to a conventional waffle raft using a multi-criteria analysis. The assessment evaluates the structural performance, cost efficiency, and sustainability using finite element modelling, life cycle cost analysis, and life cycle assessment, respectively. The structural performance of the developed prefabricated system was found to have reduced the deformation and cracking by approximately 40%. However, the cost, GHG emission, and embodied energy were higher in the prefabricated footing system due to the greater required amount of concrete and steel than that of the waffle raft. The cost difference between the two systems can be reduced to as low as 6% when prefabricated systems were installed in a highly reactive sites with large floor areas. The life cycle assessment further observed that the prefabricated footing systems consume up to 21% more energy and up to 18% more GHG emissions. These can significantly be compensated by reusing the developed prefabricated footing system, decreasing the GHG emission and energy consumption by 75–77% and 55–59% with respect to that of the waffle raft.

1. Introduction

Prefabricated construction refers to the process of building a structure through off-site manufacturing of elements under controlled conditions and then transporting and assembling it on-site. This can potentially alleviate the challenges experienced by the Architecture, Engineering and Construction (AEC) industry since prefabrication expedites the total construction period by eliminating delays due to weather impact and skilled labour shortages [1,2,3,4,5]. Prefabricated construction is also a more sustainable option for reducing the generation of waste considering the entire life cycle stages of a building, and further enables a reusable and relocatable structure [6,7,8]. Furthermore, prefabricated systems have better and more consistent quality through a controlled manufacturing process and a safer working environment [9,10,11].

Most observed prefabrication advancements have been focusing on superstructures, while the conventional cast-in-situ method, particularly for concrete structures, is still the preferred construction technique for footing systems [12]. The full potential of prefabricated construction is yet to be fully realised since most innovations have been focusing on the superstructures of buildings.

Prefabrication of footing systems requires an inter-disciplinary approach to consider critical manufacturing, structural and geotechnical aspects [13]. Important considerations include the appropriate design to account for the susceptibility of the footings of lightweight structures on clayey footings to severe damage induced by reactive soils [14,15]. Reactive soils are clayey footings, which undergo a significant change in volume due to seasonal soil water content changes. These soils experience swelling when the water level increases and shrinking when the water level decreases [16,17]. The difference in soil water level between the uncovered ground, where evaporation, transpiration and precipitation take place, and the covered ground, where impermeable footings are constructed, causes differential ground movement (Figure 1). The shrink–swell soil movement causes structural deformation through soil–structure interaction. This is crucial for lightweight structures such as houses, garages and pavements [18,19]. Repair cost amounts to billions of dollars worldwide to mitigate the damage induced by the shrinking and swelling of reactive soils [13,20,21,22,23].

Due to the absence of prefabricated footing systems for houses, a prefabricated design was developed considering the detrimental effects inflicted by the shrink–swell ground movement. The main aim of this study is to present and assess the global structural performance, cost efficiency, and sustainability of a developed prefabricated footing system using a multi-criteria analysis. The multi-criteria analysis will be based on structural performance, life cycle cost (LCC) analysis, and life cycle assessment (LCA). The developed prefabricated footing system is compared to a conventional in-situ monolithic waffle raft.

2. Methodology

The developed prefabricated footing system is assessed and compared to a conventional in-situ monolithic waffle raft. The comparison will determine which option has a more sustainable design based on three criteria, as described in Figure 2a, having better structural performance, more cost efficient, and lesser environmental impact. The developed prefabricated footing system is presented in the following section, whilst the design of the conventional waffle rafts followed the design stipulations in AS 2870-2011 by [24]. The multi-criteria analysis will evaluate the structural performance using a coupled hydro-mechanical FEM, cost analysis using LCC, and environmental impact (GHG and embodied energy) using LCA, as described in Figure 2b and discussed in the following sections.

2.1. The Developed Prefabricated Footing

The developed prefabricated footing system had a constant floor area of 225 m${}^2$ (15 by 15 m${}^2$), since the average floor area of single-detached dwellings in Australia is approximately 230 m${}^2$ [18].

Dimensional coordination is considered in the design of the developed prefabricated footing system. This manufacturing concept organises different elements independently with the goal of being connected and integrated as a whole system [25,26]. This approach does improve the assembly of a structure considering strict tolerance and improves the design flexibility of structural components. The hypothetical developed prefabricated footing systems with a 225 m${}^2$-floor area are divided into three main components to achieve dimensional coordination. The three main components are: (1) the developed dapped-end beams (DEB) and bolt connection; (2) connectors (i.e., 2-way, 3-way and 4-way connectors); and (3) slabs. The plan view of the developed prefabricated footing systems showing the three main components is presented in Figure 3. These structural elements were also designed considering the dimensional limits for handling and transportation. Through the connectors, the length of the DEB can be varied depending on the required span of the entire footing floor area achieving dimensional coordination.

Sites in Australia are classified based on the reactivity of founding soils where structures are being constructed described in Standards Australia [24]. The reactivity of soil is related to the characteristic surface movement, $y_s$, presented in

Table 1. Classification of site based on reactivity of soil and soil movement, $y_s$, as described in the Australian Standard (AS) 2870-2011 [24].

Class	Soil Footing	${\mathit{y}}_{\mathit{s}}$ (mm)
A	gravelly and sandy soil	0
S	slightly reactive silt or clay soil	0 to 20
M	moderately reactive silt or clay soil	20 to 40
H${}_1$	highly reactive clay soil	40 to 60
H${}_2$	very highly reactive clay soil	60 to 75
E	extremely reactive clay soil	greater than 75
P	filled, soft silt or clay, loose sands,	varying
	sandslip, mine subsidence, collapsing
D	areas having deep-seated soil moisture changes	-
	shall use a suffix “-D”

, calculated as:

$$y_s=\frac{1}{100}\sum _{n=1}^N{(I_{pt}\Delta \overline{u}{h}_{soil})}_n=\frac{1}{100}\sum _{n=1}^N{(\alpha I_{ss}\Delta \overline{u}{h}_{soil})}_n,$$

(1)

where $I_{pt}$ is the index for instability (in pF), $\Delta \overline{u}$ is the average change in soil suction over the thickness of the soil layer being considered (in pF), $\alpha$ is a factor to consider soil lateral restraint, $I_{ss}$ is an index for soil shrinkage (in %/pF), $h_{soil}$ is the thickness of the soil layer being considered (in m), and N is the assumed number of soil layers. The specified design of footing systems, in this specific case the hypothetical waffle rafts, will be dependent on the site classification outlining the required beam dimensions and steel reinforcements.

Performance of the prefabricated footing design for each site classification in Table 1 was compared to waffle rafts as stipulated in AS 2870.4.6 [24]. The values of unit stiffness ($EI/L$) of the monolithic waffle rafts required for each site class were first calculated using Standards Australia [24]:

$$\frac{EI}{L}=log\left[\frac{\sum \frac{B_w{D}^3}{12}}{W}\right],$$

(2)

where $B_w$ is the width of the external and internal beams (in mm), D is the depth of beams (in mm), and W is the width of a slab perpendicular to L (in m). The hypothetical waffle rafts had a constant slab thickness of 85 mm. Steel reinforcements were 3–16 mm ⊘ (3N16) for the edge beams, 1–16 mm ⊘ (1N16) for the internal beams, and 8 mm ⊘ spaced at 200 mm (SL82) for the slab mesh, as stipulated in Standards Australia [24]. Using Equation (2), $EI/L$ for each site classification was calculated and was used to determine the equivalent dimensions of the hypothetical prefabricated footing systems per site class. This will effectively compare the structural performance between the hypothetical monolithic waffle rafts and the developed prefabricated footing systems for each site class (i.e., Class M/M-D, Class H${}_1$/H${}_1$-D and H${}_2$/Class H${}_2$-D). The calculated $EI/L$ and the corresponding concrete beam dimensions are shown in

Table 2. Sections of the developed prefabricated footing system for each site class.

Class	Type	${\mathit{B}}_{\mathit{w}}$ (mm)	D (mm)	L (m)	No. of Beams	$\mathit{E}\mathit{I}/\mathit{L}$
M/	waffle raft	110	310	15	13	8.5
M-D	prefabricated	300	360	15	4	8.5
H${}_1$/	waffle raft	110	385	15	13	8.8
H${}_1$-D	prefabricated	300	450	15	4	8.8
H${}_2$/	waffle raft	110	460	15	13	9.0
H${}_2$-D	prefabricated	300	530	15	4	9.0

. The section of the developed prefabricated footing system for each site class is described in Figure 4.

2.2. Structural Performance Using a Hydromechanical Model

An advanced three-dimensional coupled hydromechanical finite element model, developed and validated by Teodosio et al. [18], Teodosio et al. [27], was used to investigate the structural performance of a prefabricated footing system. Fluid diffusion in an isotropic porous soil medium was modelled as [28,29]:

$$k_{sat}{S}^3{\psi }_w\left[\frac{\partial }{\partial x}\left(\frac{\partial h}{\partial x}\right)+\frac{\partial }{\partial y}\left(\frac{\partial h}{\partial y}\right)+\frac{\partial }{\partial z}\left(\frac{\partial h}{\partial z}\right)\right]=m_{sw}{\gamma }_w\frac{\partial h}{\partial t},$$

(3)

where $k_{sat}$ is the fully-saturated soil permeability, S is the soil saturation, ${\psi }_w$ is the pore water pressure or soil suction, h is the water potential head, $m_{sw}$ is the slope of the soil–water characteristic curve (SWCC), ${\gamma }_w$ is the unit weight of water and t is time.

The hydromechanical behaviour of reactive soils was described as [18,30,31]:

$$d{\sigma }^{\prime }=E[{\epsilon }_{es}+{\epsilon }_{ms}]=E\left[\frac{\kappa }{1+e_0}d\left(ln\left(\frac{{\sigma }^{eq}}{{\sigma }_0^{eq}}\right)\right)+\frac{1}{3}\left(d{\epsilon }_{ms}^{\prime }\frac{dS}{d{\psi }_w}d{\psi }_w\right)\right],$$

(4)

where ${\sigma }^{\prime }$ is the soil effective stress, E is the elastic constant soil tensor, ${\epsilon }_{es}$ is the soil effective stress–strain and ${\epsilon }_{ms}$ is the moisture swelling soil strain, $\kappa$ is the logarithmic soil bulk modulus, $e_0$ is the initial soil void ratio, ${\sigma }^{eq}$ is the final equivalent pressure based on the three-dimensional orthogonal stresses, and ${\sigma }_0^{eq}$ is the initial equivalent pressure.

The model used to consider concrete tensile cracking and concrete compressive crushing was the Concrete Damaged Plasticity (CDP) model [32]. The stress–strain curve shows a linear elastic relationship through ${\sigma }_{t0}$ and ${\sigma }_{c0}$ for uniaxial tension and compression. Surpassing the value of ${\sigma }_{t0}$ represent the failure stresses where the formation of cracking occurs due to the softening response from the stress–strain relationship causing strain localisation in concrete elements. Surpassing the value of ${\sigma }_{c0}$ yet less than the ultimate compression stress, ${\sigma }_{cu}$, represent the failure stresses with hardening response of the stress–strain curves. Contrarily, beyond the value of ${\sigma }_{cu}$ reflects the crushing of concrete with a softening response.

The contact element approach is employed for the interaction between reactive soils and footing systems. This method allows the soil–structure contact and separation with consideration of the tangential behaviour and normal behaviour. A more detailed discussion of the CDP model and contact elements are discussed in Dassault Systemes [33].

Stopping criterion of the developed model for all simulations was based on Standards Australia [24]. A surface pore pressure, u, was specified as a boundary condition and continuously applied to the open ground. The simulation was then terminated when a change in u below the active depth zone, $H_s$, had eventuated.

The list of parameter inputs for the numerical simulations of the developed prefabricated footing systems and the monolithic waffle rafts are presented in

Table 3. Parameter inputs for the numerical simulations of prefabricated footings and waffle rafts related to soil state, environmental factors, footing dimensions and stress condition.

Notation Soil Parameters	Parameter	Value	References
${\rho }_b$	soil density	1550 to	[34]
$\kappa$	log bulk	0.05 (swell)	[35],
	modulus	0.03 (shrink)	[30]
			[36]
			[37]
${\nu }_{soil}$	soil Poisson’s	0.45 (swell)	[38]
	ratio	0.1 (shrink)
${\epsilon }_{ms}$ vs. S	moisture-swelling	8%	[34],
	curve		[38]
$k_{sat}$	saturated permeability	1 × 10${}^{-7}$ to	[34],
		1 × 10${}^{-9}$ ms${}^{-1}$	[39]
${\psi }_w$ vs. S	sorption curve	−1 × 10${}^1$ to	[34],
		−1 × 10${}^5$ kPa	[18]
Environmental Parameters
$\Delta \overline{u}$	average suction change	1.2 pF	[24]
$H_s$	active depth zone	3.0 m	[24]
Footing and Stress Parameters
p	area load	2.5	[40],
		kN m${}^{-2}$	[41]
q	line load	6.5	[40],
		kN m${}^{-1}$	[41]
$\mu$	coefficient of	0.35 (soil-concrete)	[42],
	friction	0.4 (concrete-concrete)	[38]
$E_c$	concrete elastic	40	[43],
	modulus	GPa	[24]
${\nu }_c$	concrete Poisson’s	0.2	[43],
	ratio		[38]
$E_s$	steel elastic	450	[43],
	modulus	GPa	[27]
${\nu }_s$	steel Poisson’s	0.3	[43],
	ratio		[27]

. Details of the numerical simulations are presented in Appendix A.

2.3. Life Cycle Cost Analysis

The Life Cycle Cost (LCC) analysis was based on AC11093144 [44] to compare the overall life cycle cost of the developed prefabricated footing systems and the monolithic waffle rafts. The LCC estimates include construction and disposal costs. This study assumed that operation and occupancy costs are negligible. Present values were calculated from future costs by applying discounts. The LCC is described as

$$\mathrm{LCC}=C_C+C_D,$$

(5)

where $C_C$ is the construction cost and $C_D$ is the disposal or end-of-life cost. To effectively calculate the values of LCC for each footing system, the present value of $C_D$ can be estimated using

$$C_D^{PV}=\frac{C_D^{FC}}{{(1+d_r)}^{t_d}},$$

(6)

where $C_D^{PV}$ is the present value of the disposal cost, $C_D^{FC}$ is the future cost of the disposal cost, $d_r$ is the discount rate and $t_d$ is the structure deployment period in years.

The scale of the projects affects the calculated values of LCC. Based on past studies, the scale of a specific project determines whether a prefabricated solution will be pragmatic or uneconomical [45,46]. To consider the different scale of projects and assess the applicability of the developed prefabricated footing systems, cost analyses focusing on the comparison of material consumption and life cycle costs (LCC) are performed. These cost analyses determine the practicality of the developed prefabricated footing systems where the values of LCC are comparable or more economical to the values of the specific monolithic waffle rafts through varying the project floor area, $A_f$, and structure deployment period, $t_d$, for each site classification (i.e., Classes M-D, H${}_1$-D and H${}_2$-D). Equation for LCC can then be described as:

$$\mathrm{LCC}=\left[C_{fc}+\sum \left(C_{dc}\right)\left(A_f\right)+\sum \left(C_{ic}\right)\left(t_c\right)\right]+\frac{C_D^{FC}}{{(1+d_r)}^{t_d}},$$

(7)

where $C_{fc}$ is the fixed costs, $C_{dc}$ is the direct costs proportional to the slab floor area $A_f$, $C_{ic}$ is the indirect costs proportional to the construction period, and $t_c$ is the duration of construction.

Construction costs were derived from Rawlinsons [47], a primary reference for the AEC industry. Other resources, such as case studies, construction quotes and subcontractor estimates, were used when data were unavailable in the primary reference. Sample estimates to calculate the values of LCC for the developed prefabricated footing systems and the monolithic waffle rafts are presented in

Table 4. LCC estimate for the developed prefabricated footing with a floor area, $A_f$, of 225 m${}^2$ on a Class H${}_2$/H${}_2$-D soil deployed for $t_d$ = 50 years.

Item	Description	Qty	Unit	Rate (AUD)	Amount (AUD)
A	Site preparation
A1	Slab set-out	1	Item	600.00	600.00
A2	Mobilisation and float costs	1	Item	900.00	900.00
A3	Removal of vegetation	225	m${}^2$	2.14	481.50
	and ground leveling
A4	Site soil compaction	225	m${}^2$	1.85	416.25
A5	Installation/removal of fencing	60	Lm	42.00	2520.00
B	Formwork and reinforcement
B1	Steel reinforcement (Beam)	1.34	t	2260.00	3020.65
B2	Mesh (Slab)	0.81	t	2260.00	1836.00
B3	M20 dowel	192	Item	11.88	2280.96
B4	Plant (form release/installation)	11.63	hr	63.00	732.58
C	Concrete work
C1	Concrete mix	42.71	m${}^3$	200.00	8541.00
C2	Manufacturing plant processes	26.69	hr	60.50	1614.78
D	Delivery and installation
D1	Float of prefabricated	10,462.73	t km	0.09	941.65
	elements to site
D2	Mobilisation of crane	1	Item	500.00	500.00
D3	Crane hire (+ operator/fuel)	14.40	hr	220.00	3168.00
D4	Tradesman (+ lifting/installation)	28.8	hr	63.00	1814.40
E	Miscellaneous
E1	Concrete batching plant overheads	1	Item	3237.99	3237.99
	(20 % of construction cost)
E2	Drawings	36	Item	63.89	2300.00
E3	HSE Plan	1	Item	1500.00	1500.00
F	End of life
F1	Mobilisation of crane	2	Item	500.00	1000.00
F2	HSE Plan	1	Item	1500.00	1500.00
F3	Crane hire (+ operator/fuel)	28.80	hr	220.00	6336.00
F4	Tradesman	57.6	hr	63.00	3628.80
TOTAL LCC					53,157.06

and

Table 5. LCC estimate for the monolithic waffle raft with a floor area, $A_f$, of 225 m${}^2$ on a Class H${}_2$/H${}_2$-D soil deployed for $t_d$ = 50 years.

Item	Description	Qty	Unit	Rate (AUD)	Amount (AUD)
A	Site preparation
A1	Slab set-out	1	Item	600.00	600.00
A2	Mobilisation and float costs	1	Item	900.00	900.00
A3	Removal of vegetation	225	m${}^2$	2.14	481.50
	and ground leveling
A4	Site soil compaction	225	m${}^2$	1.85	416.25
A5	Installation (+ removal) of fencing	60	Lm	42.00	2520.00
B	Formwork and reinforcement
B1	Steel reinforcement (Beam)	0.80	t	2260.00	3020.65
B2	Mesh (Slab)	0.81	t	2260.00	1836.00
B3	Formwork	60	Lm	30.00	1800.00
B4	Waffle pods	163	No.	10.00	1630.00
B5	Tradesman (placing and tie)	17.55	hr	63.00	1106.00
B6	Tradesman (+ formwork)	17.4	hr	63.00	1096.20
B7	Labourer	6	hr	60.50	350.30
C	Concrete pour
C1	Concrete (+ delivery truck)	43.18	m${}^3$	200.00	8636.06
C2	Concrete pumping	43.18	hr	8.00	345.44
C3	Labourer (pour/vibration/finish)	47	hr	60.50	2839.57
E	End of life
E1	Demobilisation (break-up/removal)	225	m${}^2$	90.00	20,250.00
TOTAL LCC					48,244.27

, respectively.

Specific assumptions for the LCC analyses of the developed prefabricated footing systems were made. First, the site mobilisation cost was taken from Rawlinson’s Cost Guide [47] and advice from abuilding subcontractor based in Melbourne, which was assumed to be constant for both prefabricated systems and cast-in-place waffle rafts. Second, supplies of flexural, shear and mesh reinforcements were extracted from Rawlinsons [47], which include delivery costs. Third, concrete wastage is negligible since the prefabricated systems will be manufactured in a controlled environment. Lastly, labour productivity was based on Rawlinson’s Cost Guide [47] and engineering inspections were not included.

Assumptions to calculate the total values of LCC of the cast-in-situ monolithic waffle rafts include, first, the site preparation cost, and reinforcements and labour productivity were based on Rawlinsons [47], similar to the assumptions for the developed prefabricated systems. Second, the estimated cost of expanded polystyrene (EPS) for the cast-in-situ monolithic waffle rafts was quoted by a supplier based in Melbourne with an assumption of a 10% provisional cost for wastage. Third, the supply of concrete and concrete pouring cost was based on Rawlinson’s Cost Guide [47] with an added 15% as a contingency.

2.4. Life Cycle Assessment

As an environmental sustainability criterion, this study performed life cycle energy and GHG emissions for the developed prefabricated footing systems and the cast-in-situ monolithic waffle rafts. For this, this study quantified total energy consumption and GHG emissions throughout the entire life cycle stages, from product production (A1–A3), transport to the site (A4), construction (or installation, A5), and deconstruction (C1). This study does not take into account the energy consumption or GHG emissions resulting from the use of both systems. The life cycle boundaries based on EN15978 [48] for both footing systems are shown in Figure 5.

The quantification of the lifecycle energy and GHG emissions based on the system boundary shown in Figure 4 is as follows.

$$\begin{array}{cc}\hfill \mathrm{Lifecycle}\mathrm{Energy}=& \sum _{i=1}^{1}E{E}_M{\left(Comp\right)}_i+\sum _{i=1}^{1}E{E}_T{\left(Comp\right)}_i+\sum _{i=1}^{1}E{E}_C{\left(Comp\right)}_i+\hfill \\ & \sum _{i=1}^{1}E{E}_D{\left(Comp\right)}_i+\sum _{i=1}^{1}E{E}_R{\left(Comp\right)}_i,\hfill \end{array}$$

(8)

$$\begin{array}{cc}\hfill \mathrm{Lifecycle}\mathrm{GHG}=& \sum _{i=1}^{1}E{C}_M{\left(Comp\right)}_i+\sum _{i=1}^{1}E{C}_T{\left(Comp\right)}_i+\sum _{i=1}^{1}E{C}_C{\left(Comp\right)}_i+\hfill \\ & \sum _{i=1}^{1}E{C}_D{\left(Comp\right)}_i+\sum _{i=1}^{1}E{C}_R{\left(Comp\right)}_i,\hfill \end{array}$$

(9)

where, $E{E}_M$: Embodied energy of material (MJ), (Comp)i: is a particular component i = 1, 2, 3 and so forth, $E{E}_T$: Embodied energy due to transport to site, $E{E}_C$: Embodied Energy due to construction, $E{E}_D$: Embodied energy due to demolition, $E{E}_R$: Embodied energy due to reuse/recycling, and $EC$: Embodied GHG. Embodied energy and GHG data for each construction product and component (A1–A3 in Figure 4 and $E{E}_M$: in Equation (9)) were used the values in the EPiC database [49], which is open access environmental flow database for Australian building products [50], and data not provided in the EPiC database were used using the Australian national LCI (AusLCI) database.

In this study, it is assumed that a gantry crane is used in a prefabricated manufacturing plant where concrete is poured into molds and modules are loaded onto trucks (i.e., handling and transportation). The energy consumption and GHG emissions associated with the transport of products (components) to the site (A4 in Figure 4 and $E{E}_T$, $E{C}_T$ in Equations (9) and (10)) may vary depending on the type and size of transport and the distance transported. This study assumes that the transport of construction products (components) is transported by 15–18 tonne trucks, and the distance is assumed to be 50 km, which is the round-trip distance of metropolitan area in Australia.

Energy consumption and the resulting GHG emissions due to construction or assembly at the site ($E{E}_C$ & $E{C}_C$ in Equations (9) and (10)) may vary depending on the equipment operation skill level of the site worker or the site conditions. The effects of the use of equipment during construction were considered to be only the energy and carbon emissions resulting from the use of major equipment such as excavators for on-site cutting or mobile cranes for the assembly of prefabricated footing systems. In this study, the energy and carbon emission values according to the use of on-site equipment were obtained from the literature and are summarised in

Table 6. Energy and GHG emission factors used for life cycle assessment of both footing system.

Item by Life Cycle	Unit	Energy (MJ Unit${}^{-1}$)	GHG (kg CO${}_2$e Unit${}^{-1}$)	References
Construction (A)
Concrete (32 MPa)	m${}^3$	2776.00	412.00	[50]
Hot-rolled steel	kg	30.60	2.40	[50]
Gantry crane	hr	190.60	14.10	[51]
Transportation (A4, C2)
Truck (15–30 tonne)	tkm	2.71	0.203	[50]
Excavator (0.2 m${}^3$ bucket)	tkm	0.74	0.053	[51]
Mobile crane (50 ton)	hr	190.60	14.10	[51]
Construction
Excavator	MJ/hr	107.50	13.12	[51]
Concrete pump	hr	1094.30	81.40	[52]
Mobile crane	hr	190.60	14.20	[51]
End of life
Concrete demolition (C)	kg	0.007	0.00054	[50]
Recycling aggregates	kg	0.07	0.006	[50]
Recycling steel	kg	11	0.74	[50]
Benefit by recycling (D)
Recycled aggregates	kg	−0.213	−0.0169	[50]
Recycling steel	kg	−30.3	−2.4	[50]

.

In addition, the effects of the deconstruction of a building that has reached the end of its service life were applied to this study by calculating the energy consumption and GHG emission values of the concrete structure demolition, which were obtained from the literature. In this study, the benefits for the reuse and recycling of the both footing systems after deconstruction were also considered. The prefabricated footing systems can be disassembled and transported for use on another construction site. The prefabricated footing systems are assumed to be reusable after the structure deployment period ($t_d$) or after the life of the structure. On the other hand, the cast-in-place monolithic waffle rafts were assumed to be recyclable by recovering the reinforced steel rebar after deconstruction the structure.

The energy and carbon emission factors are summarised in Table 6.

3. Results and Discussion

The global performance of the developed prefabricated footing systems were investigated using a multi-criteria analysis incorporating: (1) the structural performance using a coupled hydromechanical finite element model by [18]; (2) licycle cost analysis; and (3) life cycle assessment considering energy and GHG. The results of the multi-criteria analysis are outlined below.

3.1. Structural Performance

To provide an effective comparison of the structural performance of each footing system, the substructures were exposed to varying soil movements ($y_s$) for both edge heaving scenarios (Figure 1a) and centre heaving scenarios (Figure 1b).

The developed prefabricated footing systems with the three components: (1) the developed DEB and bolt connection; (2) the connectors for dimensional coordination; and (3) the slab deck, presented in Figure 3 and Figure 4. The global performance of the developed prefabricated footing systems was compared to monolithic waffle rafts taking into account soil reactivity site classification outlined in Table 1. The dimensions of the footing systems are described in Table 2, with similar unit stiffness and constant floor area (15 m by 15 m) to effectively compare the two types of substructures.

The shrink–swell soil movement induced footing system deformation through soil–structure interaction. Results of the numerical simulations are presented in Figure 6 for swelling soils and in Figure 7 for shrinking soils.

In swelling soils, both footing systems were lifted by the heaving uncovered ground around the perimeter of the substructure (Figure 6). This soil–structure interaction response was sensible since both footing systems were designed to be constructed on-ground without any embedded beams. The developed prefabricated footing systems for each site class (i.e., M/M-D, H${}_1$/H${}_1$-D and H${}_2$/H${}_2$-D) transferred the forces and moments effectively, which led to a continuous footing deformation. Furthermore, the components of the developed prefabricated footings were functionally integrated together causing the whole system to act as simply-supported elements. Comparing both footing systems, the developed prefabricated footings were more suspended than the monolithic waffle rafts specifically when $y_s$ had higher values (Figure 6c,e). This was due to a thicker slab deck that makes the prefabricated system resist more applied area loads comparable to suspended structural elements. However, the prefabricated system requires a relatively greater amount of concrete due to the increased slab thickness.

In shrinking soils, both footing systems acted like cantilever structures shown in Figure 7. The footing systems curved outwardly due to the applied line loads at the perimeter edges. The developed prefabricated footing systems also performed homogeneously similar to the swelling scenario. This reflects the continuous transfer of forces and moments in the entire substructure. Moreover, the total footing deformation was less than that of the monolithic waffle rafts where the largest displacements were experienced at the corner of the substructures.

The developed prefabricated footing system observed generally lower structural deformation induced by the swelling soil and shrinking soil, as outlined in

Table 7. Deformation, $\Delta$, of footing systems on shrinking and swelling soil considering site classification based on soil surface characteristic movement, $y_s$ in millimetres Standards Australia [24].

Class	Scenario	Prefabricated Footing				Waffle Raft
		${\mathit{y}}_{\mathit{s}}$ = 30 (mm)	${\mathit{y}}_{\mathit{s}}$= 45 (mm)	${\mathit{y}}_{\mathit{s}}$ = 60 (mm)	${\mathit{y}}_{\mathit{s}}$ = 75 (mm)	${\mathit{y}}_{\mathit{s}}$ = 30 (mm)	${\mathit{y}}_{\mathit{s}}$ = 45 (mm)	${\mathit{y}}_{\mathit{s}}$ = 60 (mm)	${\mathit{y}}_{\mathit{s}}$ = 75 (mm)
M/	swell	23	25	-	-	25	33	-	-
M-D	shrink	25	27	-	-	25	39	-	-
H${}_1$/	swell	-	22	25	-	-	33	43	-
H${}_1$-D	shrink	-	27	38	-	-	39	51	-
H${}_2$/	swell	-	-	36	39	-	-	38	48
H${}_2$-D	shrink	-	-	32	39	-	-	42	63

. Assuming a superstructure made up of an articulated masonry veneer for all simulations, the allowable deformation (${\Delta }_{max}$) for the footing systems was 30 mm based on Standards Australia [24]. The developed prefabricated footing system obtained an acceptable deformation for both swelling and shrinking scenarios when the site classification was M/M-D ($30<y_s\le 45$ mm). Contrarily, the monolithic waffle raft on Class M/M-D site exceeded ${\Delta }_{max}$ when the soil was shrinking. Both footing systems had exceeded ${\Delta }_{max}$ when the site classifications were H${}_1$/H${}_1$-D and H${}_2$/H${}_2$-D ($y_s>45$ mm). However, the monolithic waffle rafts had significantly greater structural deformation than the developed prefabricated systems. Moreover, changing the assumption of the type of superstructure construction to cladding, instead of an articulated masonry veneer, may obtain acceptable structural deformation for the developed prefabricated systems since the value of ${\Delta }_{max}$ will change to 40 mm. This is unlikely for the monolithic waffle rafts where the values of structural deformation are more than the limit.

The monolithic waffle rafts deformed excessively specifically when experiencing shrinking ground motion. The monolithic waffle rafts, based on the numerical simulations [27], do not perform satisfactorily on shrinking soil due to insufficient top reinforcing steel bars. The main flexural reinforcements were located at the bottom portion as stipulated in Standards Australia [24]. These recommended design specifications do not fully consider the behaviour of monolithic waffle rafts when acting as double cantilevers (Figure 1b) when uncovered soils are shrinking. This shrinking scenario causes negative moment in the substructure, where the only resisting reinforcements are the meshes of slabs.

The possible reasons that caused the footing system to experience system deformation more than ${\Delta }_{max}$ are, first, the calculated $EI/L$ from the recommended design in Standards Australia [24] may have been underestimated. This recommended design was based on the analysis using the Walsh Method, an uncoupled two-dimensional approach discussed in [27]. Secondly, due to the approach used in the design standard (i.e., AS 2870-2011), consideration of the non-linear behaviour and plastic mechanisms of soil, concrete and steel were not considered. Lastly, since the approach used in the design standard was two-dimensional, the critical values at the re-entrant corners (Figure 6 and Figure 7) may have not been considered thoroughly.

The edge heave and centre heave cracking damage ($d_t$ or DAMAGET) to the developed prefabricated footings and the monolithic waffle rafts are presented in Appendix B. The prefabricated system mostly experienced the structural damage in the DEB element, which is desirable since the connection was transferring the forces and moments effectively. In the edge heave scenario, the monolithic waffle raft experienced most structural damage close to the centre of the entire footing system, which is logical since the centre is the critical area for suspended structural elements. Figure A4 shows the yielding of steel. It can be observed that the yielding of steel in the prefabricated systems and the monolithic waffle rafts were in a similar area to where the simulated cracking propagated. In the centre heave scenario, the developed prefabricated footing systems for highly reactive sites had localised structural damage around the bolt connections. This indicates that, when the line loads were applied onto the prefabricated footing systems acting as a cantilever, the bolt connection and steel reinforcements moved outwardly and yielded due to the negative moment experienced by the systems. Thus, the governing failure was around the bolt connections keeping the whole prefabricated systems intact and jointed. To mitigate this, a perimeter piling can be installed. On the other hand, extensive structural damage was observed in the monolithic waffle rafts for highly reactive sites. Severe structural damage was experienced close to the perimeter, which was around the second to third internal beams from the edge beams. This was due to the lack of reinforcements to resist the negative moment, which only relied on the mesh of the slab.

The structural performance of the developed prefabricated system using a finite element method was found to have better structural performance than the conventional waffle raft, with up to 40% reduced footing deformation. Furthermore, the prefabricated system had reduced concrete cracking due to the material layout designed for both positive and negative bending moment considering the swelling and shrinking of reactive soils.

3.2. Life Cycle Cost Analysis

The structural performance of the developed prefabricated footing systems for different site classifications (Site Classes M/M-D, H${}_1$/H${}_1$-D and H${}_2$/H${}_2$-D) was more adequate than the monolithic waffle rafts based on the numerical simulations. However, the quantity of concrete and steel materials used in the design should also be considered for an effective comparison.

Table 8. Material quantity survey of the developed prefabricated footings systems and the monolithic waffle rafts.

M/M-D Concrete	Prefabricated Footings			Waffle Rafts
	Section Vol. (m${}^3$/unit)	Number (unit)	Vol. (m${}^3$/unit)	Section Vol. (m${}^3$/unit)	Number (unit)	Vol. (m${}^3$/unit)
Beams	0.50	8	3.96	0.85	13	11.05
Slabs	28.13	1	28.13	19.13	1	19.13
Total			32.09			30.18
Steel	Diameter	Length	Mass	Diameter	Length	Mass
	(m${}^2$)	(m)	(kg)	(m${}^2$)	(m)	(kg)
Flexural	0.000491	1234.21	522	0.000113	390.00	449
Shear	0.000201	159.36	243	-	-	-
Mesh	0.000050	2250.00	625	0.000038	2250.00	625
Bolts	0.000314	57.60	16	-	-	-
Plates	0.001100	12.00	8	-	-	-
Total			1424			1074
H${}_1$/H${}_1$-D	Section vol.	Number	Vol.	Section vol.	Number	Vol.
Concrete	(m${}^3$/unit)	(unit)	(m${}^3$)	(m${}^3$/unit)	(unit)	(m${}^3$)
Beams	0.90	8	7.20	1.13	13	14.70
Slabs	28.13	1	28.13	19.13	1	19.13
Total			35.33			33.83
Steel	Diameter	Length	Mass	Diameter	Length	Mass
	(m${}^2$)	(m)	(kg)	(m${}^2$)	(m)	(kg)
Flexural	0.000491	1706.89	722	0.000113	390.00	449
Shear	0.000201	220.39	336	-	-	-
Mesh	0.000050	2250.00	813	0.000050	2250.00	813
Bolts	0.000314	57.60	16	-	-	-
Plates	0.001100	12.00	8	-	-	-
Total			1895			1262
H${}_2$/H${}_2$-D	Section vol.	Number	Vol.	Section vol.	Number	Vol.
Concrete	(m${}^3$/unit)	(unit)	(m${}^3$)	(m${}^3$/unit)	(unit)	(m${}^3$)
Beams	1.26	8	10.08	0.76/1.42	13	14.70/18.42
Slabs	28.13	1	28.13	19.13	1	19.13
Total			38.21			33.83/37.55
Steel	Diameter	Length	Mass	Diameter	Length	Mass
	(m${}^2$)	(m)	(kg)	(m${}^2$)	(m)	(kg)
Flexural	0.000491	2127.04	890	0.000201	390.00	798
Shear	0.000201	274.64	419	-	-	-
Mesh	0.000050	2250.00	625	0.000050	2250.00	625
Bolts	0.000314	57.60	16	-	-	-
Plates	0.001100	12.00	8	-	-	-
Total			1959			1423

outlines the total material quantity used for the footing systems with respect to each site classification. The amount of concrete used for the developed prefabricated system for each site class ranged from 32 to 38 m${}^3$, comparable to that of the monolithic waffle rafts which ranged from 30 to 38 m${}^3$. Most of the difference in concrete consumption was due to the thickness of the slab between the developed prefabricated systems, equal to 125 mm, and the monolithic waffle rafts, equal to 85 mm. A thicker slab of the developed prefabricated footings was necessary to prevent excessive deformation. Steel reinforcements for the prefabricated footing systems were 350 to 630 kg greater than the reinforcements in the waffle rafts. This was due to the required steel reinforcements in the D-region portions of the connections and the top flexural reinforcements to resist negative bending moments.

The values of LCC were estimated for each site using Equation (7) for the footing systems on varying soil classifications (Figure 8). The estimates show that the values of LCC for the cast-in-place monolithic waffle rafts were consistently lower than the values of LCC for the developed prefabricated footing systems. This construction cost difference was mainly due to the consumed materials, specifically for the supply of concrete and steel reinforcements. However, the percentage difference of LCC between the footing systems can be minimised. Based on Figure 8, the percentage difference of LCC between the two systems can be reduced to 6% when the prefabricated system was installed in a highly reactive site with larger $A_f$.

The cost analysis observed that dwellings having larger $A_f$ assembled on very highly reactive soil can be cost-competitive averse to cast-in-place waffle rafts.

3.3. Life Cycle Assessment: Energy and Ghg

The life cycle energy and GHG emissions of two footing systems (prefabricated footing systems and on-site waffle rafts) are presented in Figure 9. Energy consumptions and GHG emissions of the two systems are divided into four stages according to EN15978 [48]. Each of these stages includes material production (A1–A3), construction (A4–A5), end-of-life (C) and recycling/recovery (D) or Reuse (D(reuse)).

Energy consumption in the material production occupies about 75% of the total, and GHG emissions occupy 84% of the total, indicating that more than three quarters of the total energy and GHG emissions occur in the material production stage (Figure 9). Prefabricated footing systems use relatively higher amounts of concrete (approximately 2–6% higher depending on each site classification) and steel (33–50%) compared to waffle rafts (see Table 8). This causes an increase in energy and GHG emissions for product production (stage A) of the prefabricated footing systems. As shown in Figure 9, the GHG emissions of the product production (stage A) of the prefabricated footing systems is about 8–18%, and the energy consumption is about 12–21% more depending on each site classification (Figure 9).

The energy and GHG emissions due to transport of necessary materials (stage A4) are dependent on the distance from manufacturing plants to construction sites. In the prefabricated footing systems, the transport distance may be longer than the onsite waffle rafts because the material should be considered from the manufacturing plant to the prefabrication plant and to the construction site. This can cause an increase in energy consumption and corresponding GHG emissions due to transport of materials and products. Transport impact of prefabricated footing systems (A4 stage) was 2% to 13% higher (depending on each site classification) than that of on-site waffle pods systems (Figure 9). However, the impact of transportation is less than 5% of the total, and this difference is insignificant in the overall life cycle impact (Figure 9).

In terms of energy consumption and GHG emissions due to the use of equipment on the site, the prefabricated footing systems are shown to be relatively superior, having 18% to 22% less GHG emissions compared to on-site waffle pods systems (Figure 9b,d). However, the impact of energy and GHG emissions due to construction (stage A5) was found to be insignificant, contributing only 1% (GHG emissions) or 2% (energy consumption) of the total impact.

End-of-life stage (stage C) includes deconstruction, transportation and processing of waste. The impact of this stage C on overall energy and GHG emissions does not differ significantly between the two footing systems. GHG emissions at this stage was found to account for 11–12% of the total in both systems, and energy consumption was found to account for 17–18% of the total. Between the two footing systems, the prefabricated system has 17–24% more GHG emissions and 18–25% more energy consumption compared to the on-site waffle rafts. This is because the prefabricated footing systems use relatively more material compared to on-site waffle rafts (as shown in Table 8), therefore energy and GHG emissions from deconstruction and waste disposal stages are also correspondingly higher.

Recycle or recovery of material (stage D) represents benefits obtained from the recycling and recovering of waste. Recently, this stage has gained growing attention because reuse and recycling do not only reduce construction waste sent to landfills but also decrease raw materials used for new construction. In this study, this stage was analysed into two cases; the recycling of concrete and steel (represented as stage D) and the reuse of prefabricated footing systems (denoted as D(reused)) in Figure 9. Energy and GHG emissions were evaluated by dividing the case of recycling into general aggregates or recycled steel through deconstruction like concrete products, and the case of promoting reuse at other construction sites by disassembly and assembly, specifically in the case of the prefabricated system. The energy and GHG reduction effect of the recycling or reuse of each major material/product can be represented as a negative value, and the reduction effect is shown in Figure 9. GHG reduction due to the recycling of concrete and steel is effective from 27% to 35% of the total life cycle GHG emissions, and energy consumption is shown to have the effect of reducing 45% to 57% of total life cycle energy. The prefabricated footing systems have a carbon reduction effect of about 4% (Class M and H2) to a maximum of 8% (Class H1) more than that of the waffle rafts, and the energy consumption is shown to be reduced by 6% (Class M and H2) to 11% (Class H1), which is more effective than waffle rafts. As shown in Figure 9c,d, stage D was closely related to stage A and C. Since the prefabricated systems require a relatively large amount of material compared to waffle rafts, energy consumption and GHG emissions for deconstruction and waste processing were relatively higher than that of the waffle rafts (stage C in Figure 9). On the other hand, the benefit from the recycling of waste was also higher in the prefabricated footing system compared to the waffle rafts (stage D in Figure 9) because the amount of waste recycled due to disassembly also increased.

In stage D, the prefabricated footing systems can be easily disassembled due to its design for disassembly (DfD), so it can be effectively reused in another site of similar size, as well as for recycling into aggregates or recycled steel. In this case, the reduction effects can be significantly higher because reuse of the prefabricated footing systems reduces material use (stage A) at other construction sites by the same value. In Figure 9a,b, stage D (reused) of the prefabricated footing systems show the carbon and energy reduction effect obtained by reuse at other sites after disassembly. Considering this, the life cycle GHG of the prefab system is about 23% to 25% (23% for M/M-D, 24% for H1/H1-D, 25% for H2 and 23% for H2-D) compared to the waffle pod, and the reduction effect is quite large. The life cycle energy was also found to be about 41%–45% consumed by the prefabricated footing systems (41% of M/M-D, 43% for H1/H1-D, 45% for H2 and 41% for H2-D), which is less than half of the waffle rafts.

Figure 9c,d shows the contribution rate (%) of each stage to the carbon and energy consumption in the life cycle only considering recycle wastes as aggregates and steel.

The life cycle assessment observed that the prefabricated footing systems consume up to 21% more energy for very highly reactive sites, and up to 18% more GHG emissions. When considering the recycling of both footing systems, the difference was less than 10%. Considering the possibility that the prefabricated footing systems can be reused in other sites, the energy and GHG reduction effect is expected to be much greater than that just recycling concrete or steel and, in this case, the GHG is decreased by 75–77% compared to the waffle rafts (55–59% for energy consumption compared to waffle rafts). Thus, the potential energy usage and environmental impact can be significantly compensated for by reusing the prefabricated footing systems.

4. Conclusions

This study investigated the global performance of developed prefabricated footing systems using a multi-criteria analysis for structural performance using FEM, cost efficiency using LCC, and environmental impact using LCA. The developed prefabricated footing systems were compared to cast-in-place monolithic waffle rafts for Site Classes M/M-D, H${}_1$/H${}_1$-D and H${}_2$/H${}_2$-D.

The structural performance was investigated using a three-dimensional hydromechanical FEM. This obtained a more satisfactory global structural performance for the prefabricated footing systems than the traditional cast-in-place monolithic waffle rafts, having the same stiffness for each class site. The prefabricated footing system had reduced deformation and concrete cracking due to the material layout designed for both positive and negative bending moment considering the swelling and shrinking of reactive soils. However, the material consumption of the developed prefabricated footing systems is much greater than that of the cast-in-place waffle rafts. The percentage difference of LCC between the two systems can be reduced to as low as 6% when prefabricated systems were installed in a highly reactive site with larger $A_f$.

The LCA evaluation, considering only material production, transportation and deconstruction for both systems, observed the prefabricated footing systems to consume up to 21% more energy for Class H${}_2$, and up to 18% more GHG emissions (Class H${}_2$). When considering the recycling of both footing systems after the deconstruction of the building, the prefabricated footing systems still showed high energy consumption and GHG emissions; however, the difference was less than 10%. Considering the possibility that the prefabricated footing systems can be reused in other sites, the energy and GHG reduction effect is expected to be much greater than that of only recycling concrete or steel and, in this case, the GHG is decreased by 75–77% compared to the waffle rafts (55–59% for energy consumption compared to waffle rafts).

From the overall result of the multi-criteria analysis we can observe that the developed prefabricated footing systems can be more advantageous than the conventional in-situ monolithic waffle rafts when prioritising structural robustness, constructing a large house floor area in a site with a highly reactive classification, and considering potential future substructure reuse.