Optimizing Laundry for Sustainability: Balancing Washing Efficiency and Environmental Impact in the Clothing Use Phase

Abstract

The use phase of clothing contributes significantly to the overall environmental impacts due to clothing care practices. Decreasing environmental impact while maintaining washing performance in the use phase can be an effective strategy for sustainability and circularity in the textile value chain. However, existing studies on the environmental impacts of use phase usually consider limited washing conditions and neglect their impacts on washing efficiency. This study proposes a research framework that integrates the Response Surface Methodology (RSM) and Life Cycle Assessment (LCA) methodology to optimize washing parameters for better washing efficiency with less environmental impact in the clothing use phase. A series of laundry experiments were conducted to simulate household laundry, and an environmental impact assessment was conducted based on the experimental data. The optimized washing parameters were explored under eight impact categories and in terms of washing efficiency, and comparative analyses were conducted between three different washing scenarios. The results indicated that input load is the most significant factor influencing both washing efficiency and environmental impact, but with a negative correlation. The optimized washing conditions provided effective trade-offs, demonstrating notable environmental benefits through the scenario study. In the daily washing scenario with an expectation for a middle level of washing efficiency, using the optimized washing conditions can reduce the environmental impact by 80% on average compared to the high-washing-efficiency-oriented washing process and 60% on average compared to the low-environmental-impact-oriented washing process. However, for high washing efficiency demand, optimized washing conditions are less competitive due to increased washing time and detergent use. The results emphasized the importance of choosing appropriate washing parameters according to the demand for washing efficiency. Consistent environmental improvements can be achieved by changing consumer washing habits.

1. Introduction

Increasing global clothing production and consumption, along with the decrease in average clothing lifespan, have led to tremendous sustainability issues in the textile and clothing sector [1]. It is reported that the average wearing count of clothing decreased by 36% during the first 15 years after the 21st century began, while production has approximately doubled [2]. Existing Life Cycle Assessments of clothing show that the use phase significantly contributes to overall environmental impacts [3], including considerable consumption of water, energy, and chemicals, along with the release of chemicals and microfibers [4]. The energy use associated with the use phase may account for 50% to 80% of the overall clothing life cycle for clothing washed frequently [3]. The greenhouse gas emissions in the use phase are also significant and may exceed those of the production phase, depending on the washing frequency and carbon intensity of energy consumption [5]. Reducing the energy and water consumption of care practices and extending the use phase of clothing life cycles are recognized as the most effective actions for sustainability and circularity in the textile value chain, as they have great potential to prevent waste and reduce resource consumption [6,7,8].

To reduce environmental impacts in the clothing use phase, political regulations such as energy labeling are applied in the assessment of most household washing machines; meanwhile, washing efficiency must meet a certain limit value [9]. The common strategy among manufacturers is to prolong the washing time and reduce the temperature [10], as temperature is the primary factor contributing to energy consumption and may exacerbate clothing damage in the use phase [11].

However, washing frequency and parameters, which can be very influential on total environmental impact, are determined mainly by consumers [12]. Consumers may not use energy-efficient programs with long wash durations due to inconvenience, and some perceive long cycles as not truly energy-saving [13]. Thus, on the one hand, consumers may choose short cycles with higher temperatures or more detergent to meet their needs for stain removal and good hygiene performance [10]. On the other hand, if insufficient washing efficiency leads to consumer dissatisfaction after laundry, the potential for additional washing cycles may arise, which directly increases environmental impacts. Therefore, in addition to the direct effect on environmental impact, the different care practices may also affect laundry properties, resulting in indirect environmental burdens.

Washing efficiency, as an important aspect of the washing process and a major cause of consumer dissatisfaction [14,15], has been investigated in several previous studies aimed at reducing energy consumption and maintaining hygiene levels [16]. According to the Sinner Circle, washing efficiency is a result of the synergistic actions between temperature, detergent, mechanical action, and washing duration. Reducing one parameter requires compensating for the increase in others to maintain washing performance [17,18]. For example, using detergent with a higher dosage, decreasing the load, and prolonging the washing duration can compensate for a decrease in temperature. However, the alternatives may not only cause inevitable damage to fabric, thus shortening the lifespan of clothing, but also aggravate environmental impacts [19,20].

Several studies have taken the use phase into account when conducting clothing Life Cycle Assessments, but with limited washing conditions and consideration of washing efficiency [21,22,23]. A study compared the soiling removal and corresponding environmental impacts between washing at 60 °C and washing at 40 °C with a higher detergent concentration, but no effective trade-off was found with restricted conditions [24]. It is evident that a broader range of washing conditions and a more systematic experimental design are required. Another study optimized washing parameters in a washing machine using an experimental design to balance washing efficiency and energy consumption, but it lacked consideration of consumer-oriented parameters, such as load and detergent dosage [16].

The environmental impact in different categories resulting from various washing parameters, as well as the relationship between washing efficiency and environmental impact, have rarely been examined in previous studies. The Response Surface Methodology (RSM) with a Box–Behnken design (BBD), which can establish mathematical models and assess the interactions between factors and responses at a low cost and with high precision, is well-suited to exploring the relationship between washing parameters, washing efficiency, and corresponding environmental impact. This method is especially useful when minimizing or maximizing response variables and determining the optimal conditions [25].

Therefore, this study aims to explore optimized washing conditions that achieve higher washing efficiency and lower environmental impact under different washing scenarios. By integrating the Response Surface Methodology and Life Cycle Assessment methodology, the environmental impacts of various washing processes are characterized, and the relationship between washing parameters, washing efficiency, and environmental impacts is established accordingly. Two comparative analyses on environmental impact targeting different needs of washing efficiency are conducted between three different scenarios under the Life Cycle Assessment methodology.

2. Materials and Methods

As a powerful statistical tool for designing experiments exploring the mathematical correlation between responses and independent variables [26], the Response Surface Methodology was employed to design an experiment of 15 trials. The design aimed to investigate the relationship between washing parameters and washing efficiency in both short-term and long-term washing. Simultaneously, the corresponding environmental impacts of 15 washing processes were recorded and characterized through the Life Cycle Assessment methodology, following the instructions of the ISO 14040 and ISO 14044 standards [27,28]. The optimized washing conditions were selected based on the optimal responses of performance. A comparative study was conducted between different scenarios to quantify the environmental benefits of the optimized washing process. The integrated workflow is shown in Figure 1 and detailed in this section.

2.1. Material

Following the instruction of Standard ISO 6330 [29], Type I 100% Cotton ballast (Testfabrics, Inc., West Pittston, PA, USA) was used in this experiment. To test washing efficiency, the 12 cm × 12 cm EMPA standard test cloth (106) (VERSON VLIES COURCIER)—cotton fabric soiled with a mixture of IEC carbon black and mineral oil—was used in this experiment. All experiments were conducted with 3 soiled fabrics sewn to the border of the ballast, and the corresponding weight of the ballast was adjusted to meet the designed load parameter. Commercial detergent with an adapted formulation was used in this experiment, including 5–15% anionic surfactants and non-ionic surfactants, <5% soap, polycarboxylates, and no fluorescent brightener. A commercial front-loading washing machine, the Samsung “WW80K5410UX”, was used to simulate household laundry. A water meter and a power meter were installed outside the washing machine to measure the water and energy use. Water hardness was controlled at a maximum of 3.0 mmol/L, expressed as calcium carbonate, following ISO 6059 [30].

2.2. Experimental Design

2.2.1. Design of Experiment

Consumer-oriented parameters affecting washing efficiency, energy, and water consumption were considered: load, detergent concentration, temperature, and washing cycles [15,31]. According to information disclosed by the European Environment Agency and A.I.S.E. regarding residents’ laundry habits, the average load per cycle is 4 kg, while the average washing temperature is 40 °C, and only around 15% of washing is done below 20 °C or above 60 °C [32,33]. Regarding detergent, the usage amount usually varies based on the product, the degree of soiling, and the water hardness. The recommended amount of detergent used in this study was 35 mL for a 4–5 kg load washed with soft to medium–hard water, which is approximately 3000 ppm in the case of 12 L water injection during the wash stage. Considering the variation range of consumer laundry habits, the load varied from 2 kg to 6 kg, the detergent concentration varied from 1500 ppm to 4500 ppm, and the temperature varied from 20 °C to 60 °C [12,34].

To accommodate a large number of long-term wash experiments (up to 25 cycles for different parameters), a Box–Behnken Design (BBD) was used to generate the minimum number of required experimental trials in need of quadratic model fitting [35,36]. A design of three factors at three levels was employed for the experiments, as shown in

Table 1. The Box–Behnken Design for independent variables.

Trial	Temperature (${\mathit{x}}_1$, °C)	Load (${\mathit{x}}_2$, kg)	Detergent Concentration (${\mathit{x}}_3$, ppm)
1	20 (−1)	4 (0)	1500 (−1)
2	40 (0)	2 (−1)	4500 (1)
3	40 (0)	6 (1)	4500 (1)
4	60 (1)	6 (1)	3000 (0)
5	40 (0)	2 (−1)	1500 (−1)
6	20 (−1)	6 (1)	3000 (0)
7	60 (1)	4 (0)	4500 (1)
8	60 (1)	4 (0)	1500 (−1)
9	40 (0)	4 (0)	3000 (0)
10	60 (1)	2 (−1)	3000 (0)
11	40 (0)	6 (1)	1500 (−1)
12	40 (0)	4 (0)	3000 (0)
13	20 (−1)	4 (0)	4500 (1)
14	40 (0)	4 (0)	3000 (0)
15	20 (−1)	2 (−1)	3000 (0)

. The process parameters (independent variables) were temperature ($x_1$), load ($x_2$), and detergent concentration ($x_3$). Each experimental trial involved 25 washing cycles.

The standard least-squares method was used to fit the second-order polynomials, considering linear terms, square terms, and interaction terms. The equation is presented below:

$$\mathrm{y}={\beta }_0+{\sum }_{i=1}^k{\beta }_i{x}_i+{\sum }_{i=1}^k{\beta }_{ii}{x}_i^2+{\sum }_{1\le i\le j}^k{\beta }_{ij}{x}_i{x}_j+\epsilon$$

(1)

where y is the response defined with a goal of maximization or minimization; $x_i$ and $x_j$ are independent variables (the variables are temperature, load, and detergent concentration in this study); ${\beta }_0$ is the model intercept coefficient; ${\beta }_i$, ${\beta }_{ii}$, and ${\beta }_{ij}$ are the coefficients of the linear, quadratic, and interaction terms, respectively; k is the number of independent parameters (k = 3 in this study); and $\epsilon$ is the error [37].

To explore optimized conditions for various responses, the multiple responses were converted into a single one [38]. A composite desirability function was constructed and optimized through the following function; a gradient descent algorithm was used for continuous factors.

$$\mathrm{D}=d_1^{1/k}{d}_2^{1/k}\cdots d_k^{1/k}$$

(2)

where k represents the number of responses and $d_1, d_2, \dots ,d_k$ denote the individual desirability functions for k responses. In this study, the responses were the washing efficiency with the maximizing goal and environmental impacts with the minimizing goal.

The design of experiments, modeling, and optimization were all conducted using JMP Pro 17 statistical software (SAS Institute Inc., Cary, NC, USA).

2.2.2. Laundry Process

To prepare for the washing experiment, all the detergent was pre-dissolved in soft water. The ballast and test soiled fabrics were weighed to ensure that they met the required load masses and then put in the front-loading washing machine by picking the center point of each item to make sure that the loads were evenly distributed and received the same mechanical action with the same washing parameters. The washing program details are presented in

Table 2. Washing program details.

Washing Stages		Time (min)	On(s)/Off(s) Ratio	Rotation Speed (rpm)	Water Injection Amount (L)
Main wash stage	Water injection and heating	6	5:5	30	12
	Main wash	20	18:4	50
	Spin	5	-	1400
Rinse stage	Water injection	1.5 × 2	5:5	30	29.5
	Rinse × 2	4.5 × 2	10:6	40
	Spin (between rinses)	5	-	1400
Dehydration stage		14	-	1400	-
Total		62	-	-	41.5

.

The water meter and power meter monitored and collected water and energy consumption non-intrusively while performing the different experiment trials. After each washing cycle, all input fabrics were weighed to measure residual moisture and then spread on a horizontal drying rack in an ambient environment for 24 h before measurement [29]. The conditioning atmosphere used for textile specimen balancing had a temperature of 20.0 ± 2.0 °C and a humidity of 65.0 ± 4.0% [39]. The wastewater information was calculated by subtracting the residual moisture from the total water consumption.

Table 3. Energy consumption, residual moisture, and wastewater information of 15 designed washing processes.

Trial	Energy Consumption (kWh)	Residual Moisture (kg)	Wastewater (L)
1	0.462	1.69	39.81
2	0.829	0.69	40.81
3	0.534	2.40	39.1
4	0.133	2.40	39.1
5	0.815	0.69	40.81
6	0.828	2.40	39.1
7	0.464	1.69	39.81
8	0.778	1.69	39.81
9	0.441	1.69	39.81
10	0.464	0.69	40.81
11	0.121	2.40	39.1
12	0.464	1.69	39.81
13	0.113	1.69	39.81
14	0.118	1.69	39.81
15	0.428	0.69	40.81

presents the detailed energy consumption, residual moisture, and wastewater information.

2.2.3. Evaluation of Washing Efficiency

To analyze the washing efficiency and unevenness under different washing parameters, color measurements were carried out on the soiled fabrics and ballast before and after washing by reflectance spectrophotometry, considering a 10 observer under CIE standard D65 illuminant, as specified. The measurements were performed before the first washing cycle and after 1, 3, 5, 10, 15, 20, and 25 washing cycles. Color measurements were taken with four layers of the same washed soiled fabric samples as backing for the measured piece, as well as the attachment area on the ballast. Every sample was measured twice on both sides at the positions indicated in Figure 2. The average value of the four readings was used in the subsequent evaluation.

In this study, washing efficiency was evaluated as the color difference between the soiled fabric and the cotton ballast’s background before and after washing [40]. The calculation is presented below.

$$∆E_{CMC}=\sqrt{{\left(∆L/l{S}_l\right)}^2+{\left(∆C/c{S}_c\right)}^2+{\left(∆H/S_h\right)}^2}$$

(3)

$$\mathrm{D}=\left({∆E}_{CMCO}-{∆E}_{CMCF}\right)/{∆E}_{CMCO}\times 100\%$$

(4)

where $\mathrm{D}$ is the washing efficiency; ${∆E}_{CMCO}$ and ${∆E}_{CMCF}$ are the original and final noticeability of the soiled fabric and cotton ballast calculated through Equation (3) based on the CIE L*c*h* color model (an alternative representation of L*a*b* coordinates).

2.3. Life Cycle Assessment

2.3.1. Scope and System Boundaries

The laundry process was the stage considered from a gate-to-gate perspective in this Life Cycle Assessment. The environmental impacts of the use phase were mainly attributed to two aspects: (1) the inputs from the water, detergent, and washing machine used to realize the washing process and the energy consumption required to perform mechanical actions and heating within the washing machine; (2) the wastewater output, carrying detergents and microfibers, and the water emission from the residual water on the fabric after washing. Two different functional units were considered in this study. FU.a was the washing of 1 kg of clothing for 1 cycle, which may result in different washing efficiency under different washing parameters. FU.b was the washing of 1 kg of clothing to achieve a certain washing efficiency, such as the highest washing efficiency or a moderate washing efficiency after 1 cycle in this experiment, which points to a high-level or a middle-level washing efficiency expectation, respectively. Under FU.b, different cycles may be required to achieve a certain washing efficiency using different washing parameters. According to each functional unit, two comparative analyses were conducted between different scenarios.

2.3.2. Life Cycle Inventory

The main input and output data were collected during the washing process in the laboratory. The energy consumption, residual moisture, and wastewater emission information for each trial is exhibited in Table 3. It was assumed that the washing machine would serve a lifetime of 12 years, perform 208 cycles per year, and run 2.15 h per cycle on average [41,42,43,44]. The detergent inventory data were modeled based on the report provided by the supplier, including the ingredient composition; consumption of electricity and heat in the production; and transport between the raw material supplier, the factory, the distribution center, and the consumer. Based on these data, the Ecoinvent 3.10 database was used to populate the necessary dataset for Europe to the greatest extent possible.

2.3.3. Impact Assessment Method

For the environmental impact assessment, the Life Cycle Assessment tool Simapro 9.6 (PRé Sustainability, Amersfoort, The Netherlands) was employed with the European environmental footprint methods (EF 3.1, 2023). The EF 3.1 impact assessment methods are built upon established international practices and standards by the European Commission and aim to help consumers make informed choices based on reliable and verifiable information [45]. Sixteen midpoint impact indicators are covered, including climate change; ozone depletion; ionizing radiation; photochemical ozone formation; human toxicity, non-cancer; human toxicity, cancer; acidification; particulate matter; eutrophication, freshwater; eutrophication, marine; eutrophication, terrestrial; ecotoxicity, freshwater; land use; water use; resource use, fossils; and resource use, minerals and metals [46]. Normalization was performed for the contribution analysis, producing a single score by weighting each category according to the same units and summing them for further comparisons between different washing processes [47].

Based on the inventory robustness, reliability, and recommendation level of the EF impact assessment, 8 impact categories were chosen to explore optimized conditions, including climate change; particulate matter; eutrophication, freshwater; ionizing radiation; land use; photochemical ozone formation; resource use, minerals and metals; and water use [48]. Quantified environmental impacts with a minimizing goal (the best value is the smallest possible) and washing efficiency with a maximizing goal (the best value is the largest possible) were the responses for multiple optimization.

3. Results and Discussion

3.1. Results for Washing Efficiency

The results for the washing efficiency after each cycle are presented in Figure 3. The washing efficiency of each washing trial shows a rapid increase in the first 10 cycles, followed by a progressive slowdown over cycles, and finally leads to a plateau in the washing efficiency profile. Due to the different physicochemical and mechanical effects on cleaning, 15 washing trials present very different washing efficiency [49]. The maximum washing efficiency after the first cycle is 36.46% in Trial 10 (60 °C temperature, 2 kg load, 3000 ppm detergent concentration), which exceeds the minimum washing efficiency (8.41% from Trial 4) by a factor of four. Among the washing trials examined during the washing experiment, Trial 10 has the highest washing efficiency among all washing cycles, while Trial 6 (20 °C, 6 kg, 3000 ppm) has the lowest washing efficiency, except for the first cycle. Trial 6 needed to be performed with 10 cycles to obtain the same washing efficiency as Trial 10 with 1 cycle. As the washing cycles increased, the difference between the different trials gradually decreased. The maximum difference in washing efficiency under different washing parameters reached 28.05% after the first washing cycle, narrowed to 23.02% after 10 washing cycles, and eventually fell to 15.35% after 25 washing cycles. The big difference in washing efficiency in 15 trials revealed the importance of washing parameters in daily laundry.

Linear terms, square terms, and interaction terms were considered to fit the second-order polynomial models to predict washing efficiency after each cycle. A summary of the fitted second-order polynomial model obtained from the experiment results and analysis of variance (ANOVA) is presented in

Table 4. ANOVA of the fitting results for the prediction of washing efficiency.

Source	DF	WE-1 (%)		WE-3 (%)		WE-5 (%)		WE-10 (%)		WE-15 (%)		WE-20 (%)		WE-25 (%)
		CE	p Value	CE	p Value	CE	p Value	CE	p Value	CE	p Value	CE	p Value	CE	p Value
Model	9	21.72	0.0184	36.53	0.0003	42.6	0.0046	50.82	0.0031	54.94	0.0004	57.33	0.0003	59.26	0.0004
$x_1$	1	1.83	0.1845	2.88	0.0025	2.74	0.0198	3.89	0.002	3.76	0.0001	3.31	0.0001	2.98	0.0002
$x_2$	1	−8.87	0.0007	−9.92	<0.0001	−8.24	0.0002	−6.65	0.0002	−5.44	<0.0001	−5	<0.0001	−4.69	<0.0001
$x_3$	1	0.63	0.6206	1.18	0.070	1.98	0.0585	1.72	0.0482	1.27	0.0167	1.35	0.0069	1.2	0.0120
$x_1{x}_2$	1	−3.27	0.1102	−1.27	0.1427	−0.79	0.5205	1.86	0.1034	0.88	0.1448	0.27	0.5618	−0.04	0.9331
$x_1{x}_3$	1	3.64	0.0833	2.08	0.0355	1.85	0.1678	1.47	0.1754	1.32	0.0480	0.83	0.1119	0.91	0.0922
$x_2{x}_3$	1	−2.02	0.2848	−1.69	0.0678	−2.25	0.1072	−0.3	0.7645	−0.49	0.3802	−0.11	0.8049	0.36	0.4538
$x_1^2$	1	−0.54	0.7723	−1.41	0.1220	−1.18	0.3677	−0.86	0.4166	−1.72	0.0227	−1.9	0.0082	−1.86	0.0096
$x_2^2$	1	0.5	0.7856	0.01	0.9904	0.06	0.9588	0	0.9988	−0.26	0.6419	−0.86	0.1151	−1.27	0.0390
$x_3^2$	1	1.93	0.3219	0.57	0.4889	−0.83	0.5179	−1.21	0.2681	−1.15	0.0818	−0.83	0.1242	−1.01	0.0778
$R^2$		0.93269		0.98843		0.962635		0.968058		0.986912		0.988587		0.986515
Adj. $R^2$		0.811531		0.967605		0.895377		0.910561		0.963354		0.968043		0.962241
RMSE		3.370571		1.458411		2.294219		1.869406		1.014714		0.862348		0.878511
Mean		22.73169		36.08056		41.56447		49.71195		53.27337		55.41865		57.04911

Note: $x_1$—temperature; $x_2$—load; $x_3$—detergent concentration; WE—washing efficiency; DF—degree of freedom; CE—coefficient; RMSE—root-mean-square deviation. WE-n is the washing efficiency after n cycles.

. The prediction expression of washing efficiency after each cycle can be obtained with the model terms and the corresponding coefficient values. The ANOVA results indicate that the models are all significant (p < 0.05). In the results, the adjusted $R^2$ of a model is noticeably smaller than the $R^2$ after 1, 5, and 10 cycles and very close after 3, 15, 20, and 25 cycles. It is assumed that due to the size of the soiled fabric and the side attachment method, flexing may occur during washing, causing creases to the soiled fabric. Lighter lines were observed on the soiled fabric after the first cycle due to the higher friction rate of crease lines with other fabrics, which increased the variability in WE between the three soiled fabrics. As the cycles increased, the influence of crease lines gradually decreased. The color difference between the crease lines and the normal surface was reduced, and new crease lines occurred inevitably after a new round of washing.

From the p value for each term, we find that the load ($x_2$) is always a significantly convincing factor for WE and has a negative coefficient in both short-term and long-term washing. In the washing machine used in the experiment, a 2 kg load occupied around one-third of the capacity. With a rotation speed of 40–50 rpm, there was more chance of friction between fabrics and an increase in falling and rotating movements during washing [50]. However, soil removal is not only the result of mechanical force, it also combines synergistic actions between the temperature, detergent, and washing duration [10]. The results indicate that the variables are not equally significant, and the significance changes as the washing cycles increase. From three cycles, the temperature starts to show the significance of residual soiling, which is sensitive to mechanical actions. As the cycles increase, detergent concentration and the quadratic effects of temperature can be observed to be significant for washing efficiency at 20 and 25 cycles. The interaction profiles in Figure 4 also show that, at a relatively lower temperature, the washing efficiency is similar regardless of whether the detergent concentration is high or low. However, at relatively high washing temperatures, high detergent concentrations often lead to higher washing efficiency.

It is assumed that while the solid soiling between yarns is removed because of the effective exposure to friction and hydrodynamic forces, the residual soiling in the deep locations between yarns or intra-yarns is less accessible [40], such that more support in terms of temperature and the interaction of detergent and temperature might be needed to dissolve oily components in the residual soiling [51].

3.2. Environmental Impact Assessment

Table 5. Environmental impact assessment of 15 washing processes.

Damage Category	Unit	Washing Trials
		1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
Acidification	mol H+ eq	0.0003	0.0009	0.0003	0.0004	0.0009	0.0002	0.0006	0.0006	0.0004	0.0012	0.0003	0.0004	0.0003	0.0004	0.0006
Climate change	kg CO2 eq	0.0396	0.1505	0.0519	0.0689	0.1488	0.0307	0.107	0.0985	0.0733	0.1985	0.0444	0.0733	0.0499	0.0733	0.089
Ecotoxicity, freshwater	CTUe	1.0621	3.4879	1.1453	1.1839	3.2922	1.026	1.8536	1.7241	1.6673	3.5921	1.0512	1.6673	1.6175	1.6673	3.1397
Particulate matter	disease inc.	2 × 10−9	8 × 10−9	3 × 10−9	3 × 10−9	7 × 10−9	2 × 10−9	4 × 10−9	4 × 10−9	4 × 10−9	8 × 10−9	2 × 10−9	4 × 10−9	3 × 10−9	4 × 10−9	6 × 10−9
Eutrophication, marine	kg N eq	0.0002	0.0005	0.0002	0.0002	0.0005	0.0001	0.0003	0.0003	0.0002	0.0005	0.0001	0.0002	0.0002	0.0002	0.0004
Eutrophication, freshwater	kg P eq	4 × 10−5	2 × 10−4	5 × 10−5	7 × 10−5	2 × 10−4	3 × 10−5	1 × 10−4	1 × 10−4	7 × 10−5	2 × 10−4	4 × 10−5	7 × 10−5	5 × 10−5	7 × 10−5	1 × 10−4
Eutrophication, terrestrial	mol N eq	0.0004	0.0016	0.0005	0.0007	0.0014	0.0003	0.001	0.0009	0.0007	0.0019	0.0004	0.0007	0.0006	0.0007	0.001
Human toxicity, cancer	CTUh	5 × 10−10	1 × 10−9	4 × 10−10	5 × 10−10	1 × 10−9	4 × 10−10	7 × 10−10	6 × 10−10	6 × 10−10	1 × 10−9	4 × 10−10	6 × 10−10	6 × 10−10	6 × 10−10	1 × 10−9
Human toxicity, non-cancer	CTUh	2 × 10−9	4 × 10−9	1 × 10−9	1 × 10−9	4 × 10−9	1 × 10−9	3 × 10−9	3 × 10−9	2 × 10−9	5 × 10−9	1 × 10−9	2 × 10−9	2 × 10−9	2 × 10−9	3 × 10−9
Ionizing radiation	kBq U-235 eq	0.0097	0.0527	0.0187	0.0315	0.063	0.0071	0.0467	0.047	0.028	0.0891	0.0177	0.028	0.0102	0.028	0.0191
Land use	Pt	0.2529	1.1014	0.3762	0.4252	0.8632	0.2286	0.7142	0.5557	0.4835	1.234	0.2609	0.4835	0.4201	0.4835	0.6705
Ozone depletion	kg CFC11 eq	2 × 10−9	5 × 10−9	2 × 10−9	2 × 10−9	5 × 10−9	1 × 10−9	3 × 10−9	3 × 10−9	3 × 10−9	6 × 10−9	2 × 10−9	3 × 10−9	2 × 10−9	3 × 10−9	4 × 10−9
Photochemical ozone formation	kg NMVOC eq	0.0001	0.0005	0.0002	0.0002	0.0005	0.0001	0.0003	0.0003	0.0002	0.0006	0.0001	0.0002	0.0002	0.0002	0.0003
Resource use, fossils	MJ	0.6758	2.8835	1.0031	1.4131	2.9508	0.5292	2.1777	2.0343	1.4251	4.0487	0.8642	1.4251	0.8559	1.4251	1.5152
Resource use, minerals and metals	kg Sb eq	1 × 10−6	3 × 10−6	1 × 10−6	1 × 10−6	3 × 10−6	7 × 10−7	2 × 10−6	2 × 10−6	1 × 10−6	4 × 10−6	9 × 10−7	1 × 10−6	1 × 10−6	1 × 10−6	2 × 10−6
Water use	m3 depriv.	0.0963	0.1992	0.0792	0.0828	0.1916	0.0719	0.1195	0.1135	0.108	0.2094	0.0746	0.108	0.1031	0.108	0.178

presents the environmental impacts of 15 different washing trials under FU.a, washing 1 kg of clothing for one cycle. Among the washing parameters examined during the washing experiment, Trial 10 (60 °C temperature, 2 kg load, 3000 ppm detergent concentration) has the highest environmental impacts among all the impact categories, followed by Trial 2 (40 °C, 2 kg, 4500 ppm) and Trial 5 (40 °C, 2 kg, 1500 ppm), which all have a low input load. The lowest impact is from Trial 6 (20 °C, 6 kg, 3000 ppm) with the lowest washing temperature and the highest input load.

Figure 5 graphically shows the single score results of different washing processes and the relative contributions to environmental impacts. Depending on the different washing parameters, the breakdown of environmental impacts varies. The electricity use ranges from 10% to 45% of the environmental impact, while water consumption ranges from 22% to 37%, and the washing machine accounts for 17% to 29%. Detergent generally contributes the least to the total environmental impact, except in Trials 13 and 15, which involved a low washing temperature and relatively high detergent inputs. Both the environmental impact assessments and single score results for 15 washing trials indicate significant variations under different washing parameters, which are especially noticeable under large load and temperature differences. Therefore, it is necessary to examine the relationship between washing parameters and environmental impacts and optimize the parameters.

While performing ANOVA analysis for eight chosen impact categories (

Table 6. Effect summary of environmental impacts.

Climate Change		Particulate Matter		Eutrophication, Freshwater		Ionizing Radiation
Source	p Value	Source	p Value	Source	p Value	Source	p Value
$x_2$	<0.0001	$x_2$	<0.0001	$x_2$	<0.0001	$x_1$	<0.0001
$x_1$	<0.0001	$x_2^2$	<0.0001	$x_1$	<0.0001	$x_2$	<0.0001
$x_2^2$	0.0005	$x_1$	<0.0001	$x_2^2$	0.0002	$x_1{x}_2$	0.0021
$x_1{x}_2$	0.0016	$x_3$	0.0002	$x_1{x}_2$	0.0012	$x_2^2$	0.0062
$x_3$	0.1439	$x_1{x}_2$	0.0010	$x_3$	0.0629	$x_2{x}_3$	0.2060
$x_2{x}_3$	0.6327	$x_2{x}_3$	0.0411	$x_1{x}_3$	0.6067	$x_3$	0.4532
$x_3^2$	0.6899	$x_1^2$	0.6863	$x_1^2$	0.6468	$x_3^2$	0.6910
$x_1^2$	0.7868	$x_1{x}_3$	0.6958	$x_2{x}_3$	0.6811	$x_1^2$	0.8158
$x_1{x}_3$	0.8804	$x_3^2$	0.7423	$x_3^2$	0.7985	$x_1{x}_3$	0.9235
Land Use		Photochemical Ozone Formation		Resource Use, Minerals and Metals		Water Use
Source	p Value	Source	p Value	Source	p Value	Source	p Value
$x_2$	<0.0001	$x_2$	<0.0001	$x_2$	<0.0001	$x_2$	<0.0001
$x_1$	<0.0001	$x_1$	<0.0001	$x_1$	<0.0001	$x_2^2$	<0.0001
$x_2^2$	0.0001	$x_2^2$	0.0002	$x_2^2$	0.0001	$x_1$	<0.0001
$x_3$	0.0004	$x_1{x}_2$	0.0014	$x_1{x}_2$	0.0018	$x_1{x}_2$	0.0012
$x_1{x}_2$	0.0014	$x_3$	0.0109	$x_2{x}_3$	0.3815	$x_3$	0.0024
$x_2{x}_3$	0.0865	$x_3^2$	0.6876	$x_3$	0.6200	$x_2{x}_3$	0.3654
$x_3^2$	0.6804	$x_2{x}_3$	0.7479	$x_3^2$	0.6906	$x_3^2$	0.6980
$x_1^2$	0.7856	$x_1^2$	0.7725	$x_1^2$	0.7989	$x_1^2$	0.7493
$x_1{x}_3$	0.8849	$x_1{x}_3$	0.8602	$x_1{x}_3$	0.8984	$x_1{x}_3$	0.8181

Note: $x_1$—Temperature; $x_2$—Load; $x_3$—Detergent concentration.

), it was found that load dominates the impact factors for all the impact categories except for ionizing radiation (dominated by temperature). In addition to this, temperature, the quadratic effects of load, and the interaction effect of load and temperature also show significance among all the impact categories. The results are consistent with a previous study assessing the energy performance of European laundry machines, which found that energy use efficiency depends on load size and temperature and that water use efficiency benefits from high load size [52].

3.3. Optimized Conditions

To explore optimized conditions for higher washing efficiency and less environmental impact across different categories, the composite desirability function was constructed and optimized, as described in Section 2.2.1, following Function 2. The washing efficiency and the environmental impact for each impact category under FU.a, washing 1 kg of clothing for one cycle, were used as two responses. By simultaneously maximizing washing efficiency and minimizing the environmental impact, the desirability function balances two competing objectives and adjusts factors to maximize the desirability results [53]. The optimized conditions are presented in Figure 6. In the first three washing cycles, the optimized temperature and detergent concentration conditions for eight impact categories almost all remain at consistently low levels, at around 20 °C and 1500 ppm, respectively. In contrast, the optimal condition for load varies widely in different impact categories, ranging from 2 kg to 4.97 kg. For water use, the optimized load is the highest, as it considers the contribution to the depletion of available water. A washing machine that can adjust water intake according to the input load is beneficial for this. For ionizing radiation, the optimized load remains the lowest, since electricity is the main contributor to this category.

Starting from five washing cycles, the optimized conditions for temperature and detergent concentration change significantly. Starting from 10 washing cycles, the growth of optimal washing efficiency gradually decelerates over cycles. After 15 washing cycles, the optimized conditions start to exhibit a trend toward relatively high load levels (around 4 kg to 5 kg), accompanied by elevated temperature and detergent concentration. Under bi-objective functions, the increase in load contributes greatly to the decrease in environmental impact; meanwhile, the increase in temperature and detergent concentration, with their interaction, partially compensates for the reduction in washing efficiency associated with high load levels.

3.4. Scenario Study

Combining the washing efficiency results and environmental impact results, it is clear that the Trial 10 washing process (60 °C temperature, 2 kg load, 3000 ppm detergent concentration), with the highest washing efficiency over cycles, also causes the worst environmental impact. The Trial 6 washing process (20 °C, 6 kg, 3000 ppm), with the lowest washing efficiency from the three washing cycles, also has the lowest environmental impact. In order to quantify the potential benefits of the optimized washing process, three scenarios were developed. The first scenario (S1) represents the washing process with the maximum desirability for high washing efficiency (the Trial 10 washing process). The second scenario (S2) represents the washing process with the maximum desirability for low environmental impact (the Trial 6 washing process). The third scenario (S3) represents the washing process under optimized conditions for higher washing efficiency and less environmental impact.

3.4.1. Comparative Analysis I

The first comparative analysis of the three scenarios is depicted in Figure 7 and focuses on comparing the environmental impacts of washing 1 kg clothing for one cycle (FU.a). Considering the growth rate of washing efficiency and the changes in optimized conditions across cycles, the comparison of washing for 1, 5, and 15 cycles under three scenarios is also considered.

The results show that optimized washing conditions (S3) have consistent environmental benefits across all impact categories compared to S1 and better washing efficiency compared to S2. For climate change, after one washing cycle, compared to S1, S3 has a nearly 67% washing efficiency but only causes 19% of climate change; compared to S2, S3 has 1.8 times the washing efficiency but only causes 20% more climate change. In the case of particulate matter and land use, after one washing cycle, it is possible that with minimal increased environmental impact (within 3%), the washing efficiency can be improved by 1.7 times or more.

At five washing cycles, the optimization of washing efficiency in S3 is more obvious, as it can achieve a washing efficiency above 80% compared to S1 but only causes 40% to 50% of the environmental impact. Compared to S2, the washing efficiency is improved by 45% to 55%, but the environmental impact is more than double, except for the water use category. This is the consequence of the cumulative effect and changes in optimized conditions across washing cycles.

At 15 washing cycles, S3 has around 90% washing efficiency compared to S1; meanwhile, it only has around 30% greater washing efficiency compared to S2. This is because, starting from 10 washing cycles, the improvement in washing efficiency slows down over cycles, and the difference in washing efficiency between S1 and S2 also gradually narrows.

In Figure 7, it can also be observed that the washing efficiency of S1 after only one washing cycle exceeds that of S2 after five washing cycles. However, some environmental impacts of S1 after one washing cycle are less than those of S2 after five washing cycles, such as in the resource use, minerals and metals, and water use categories. A similar situation occurs between S1 after 5 washing cycles and S3 after 15 washing cycles. The quantitative visualization of the environmental impacts and washing efficiency clearly indicates that the better washing conditions vary at different washing efficiency levels. This is primarily due to the fact that the significant factors and their degree of effect on washing efficiency change across washing cycles. Meanwhile, the effect of washing parameters on the environmental impacts of different categories remains constant. To meet varying demands for washing efficiency with less environmental impact, different choices need to be made. Therefore, it is important to compare the environmental impact of each scenario under different expectations for washing efficiency and to identify better choices.

3.4.2. Comparative Analysis II

The second comparative study aimed to compare the environmental impacts of washing 1 kg of clothing at a certain washing efficiency (FU.b). S1 achieves a washing efficiency of 36.46% after a single washing cycle, which represents a high level of washing efficiency in this study; however, S2 requires 10 washing cycles (36.59%, shown in Figure 3), and S3 requires 3 washing cycles (36.32%, shown in Figure 6) to reach this level of washing efficiency. Under optimized conditions, S3 is able to achieve a washing efficiency of 22.67% after a single washing cycle, which is also a middle-level washing efficiency in all trials tested, while S2 requires three washing cycles (22.54%, shown in Figure 3) to reach this level of washing efficiency. Therefore, in the second comparative analysis, the function unit is adjusted to washing 1 kg of clothing to achieve a washing efficiency above a certain value, 36% and 22%, respectively. S1 is considered the reference scenario; the potential benefits and deterioration of environmental impacts in response to two different washing goals are presented in Figure 8.

Based on the results, it is evident that under different washing goals, the potential environmental impacts of each scenario also change. While having a high-level washing efficiency expectation to deal with heavy soiling (shown in Figure 8a), S2 is the process that causes the worst environmental impacts, except for the ionizing radiation category. This is primarily due to the large repetition cycles S2 needs to achieve a high level of washing efficiency. However, benefiting from the low washing temperature and high load of S2, the electricity use is significantly lower compared to S1 for a single washing cycle. This is particularly evident in the ionizing radiation impacts when electricity is supplied by a mixed source containing nuclear power [54]. Similarly, the high number of washing cycles S3 needs makes the optimized washing parameters also seem uncompetitive in achieving a high level of washing efficiency. Compared to S1, S3 reduces environmental impact by around 10% in some impact categories, such as climate change and eutrophication, freshwater. However, in particulate matter; resource use, minerals and metals; and water use, the environmental impact of S3 increased by 15%, 49%, and 69%, respectively, compared to S1. Considering time-consumption and convenience for the consumer, S1 is a better choice when a high level of washing efficiency is needed.

When expecting a middle level of washing efficiency (shown in Figure 8b), S3 shows a great advantage in environmental benefits. S3, with optimized washing parameters of a low temperature (20 °C), low detergent concentration (1500 ppm), and relatively high load (4.15 kg on average, around two-thirds of the washing machine capacity), is undoubtedly the best choice to meet daily washing needs.

The comparative analysis between three scenarios oriented towards high washing efficiency, low environmental impact, and optimized washing conditions offers valuable insights into the variability in overall environmental impact and washing efficiency under different washing parameters. It highlights the importance of choosing appropriate washing parameters according to the degree of soiling, especially since this is an overlooked aspect during the washing process for most consumers [15]. Moreover, the potential environmental benefits of changing consumer habits are huge, as the washing process will be repeated in daily household washing.

4. Conclusions

This study successfully explored the optimized washing conditions for better washing efficiency with less environmental impact in the clothing use phase by integrating the Response Surface Methodology (RSM) and Life Cycle Assessment (LCA) methodology. The results highlighted the significant influence of load for both washing efficiency and environmental impact in the washing process, but with a negative correlation. The optimized conditions identified through the RSM provide the most effective trade-offs between the two, and the environmental benefits of eight impact categories were quantified through LCA.

Scenario study further confirmed the environmental potential of optimized conditions in all eight impact categories. In the daily washing scenario with a need for a middle level of washing efficiency, using the optimized washing conditions with a low temperature (20 °C), low detergent concentration (1500 ppm), and relatively high load can reduce the environmental impact by 80% on average compared to the high-washing-efficiency-oriented washing process and 60% on average compared to the low-environmental-impact-oriented washing process. However, when considering the need for a high level of washing efficiency, optimized washing conditions are no longer competitive compared to the high-washing-efficiency-oriented washing process for longer washing times and increased detergent use.

Therefore, it is crucial to select suitable washing parameters based on the desired level of washing efficiency. Technical advances have been made by regulators and manufacturers to save energy by decreasing the temperature of pre-set washing programs. Consumer washing habits can further maximize the potential for reducing environmental impacts during the washing process. Adjusting the input load according to the degree of soiling could be the easiest and most effective method. The best conditions and their environmental performance vary across impact categories. For regions where climate change; eutrophication, freshwater; and ionizing radiation are critical concerns, washing with approximately two-thirds of the machine’s capacity at a low temperature and with a low detergent dosage consistently represents the best practice.

Scenarios requiring higher washing efficiency were not included in the comparative analysis since they exceeded the scope of general household washing. This study is limited by the single soiling and substrate type, detergent, and ballast component used in the experiment, as these variables may have affected the washing efficiency results. Environmental impact results will also have been influenced by the different detergent formulas and washing machines with varying efficiency levels used in the consumer washing scenarios. Additionally, the washing process inevitably results in fabric damage, ultimately affecting the lifespan of clothing, which should also be considered in future work.

Nonetheless, this study demonstrated the effectiveness of the integrated research framework and revealed notable potential environmental benefits from changing washing parameters without sacrificing washing efficiency. This proposed framework can also be applied in future work to explore optimal conditions for various washing properties with less environmental impact. Providing clear recommendations on clothing care practices to consumers is a key action to promote sustainability and circularity in the textile value chain, which will also enable consumers to care for clothing more effectively [7].