Assessing the benefits of urban consolidation centres: an overview based on a systematic literature review

ABSTRACT

Urban consolidation centres (UCC) have often been highlighted as a solution to reducing freight vehicle kilometres, emissions, and congestion in urban areas. However, previous studies have presented vastly different results regarding the environmental and social benefits when UCCs are implemented. Therefore, this study aims to provide an overview of research on the sustainability assessment of UCCs, to describe dominant themes, and identify why assessments differ. A systematic literature review approach employing a content analysis was used to create the overview and identify the dominant themes in the quantification of the sustainability benefits of UCCs. As a complement, a cross-case analysis was applied to compare the results and to identify underlying differences between the studies. The content analysis revealed three dominant themes, relating to: (i) modelling aspects, (ii) different UCC set-ups, and (iii) the different performance measurements applied. Furthermore, improved consolidation is often described as the largest environmental benefit of implementing UCCs but our results show that the largest benefit can be found in switching to more environmentally friendly vehicles. However, the cross-case analysis revealed difficulties in determining the benefits of implementing UCCs because the assessment of benefits differ vastly between studies. These differences can be explained by the different scope of the system and whether or not other measures were implemented alongside a UCC. This review also highlights seven important gaps in the research that can be used to guide future research, such as a lack of methodological diversity, since most studies employ mathematical modelling, as well as a lack of transparency regarding input and output data. This is a barrier when evaluating the benefits of introducing UCCs.

KEYWORDS:

• City logistics
• green logistics
• performance indicators
• closed loop supply chains
• urban transhipment
• city hub

1. Introduction

Urban logistics is associated with negative effects, including pollution, congestion, and noise (Goetz, 2019; Johansson & Björklund, 2017). While freight deliveries create inefficiencies and negatively affect the urban environment, they are an absolute necessity in today’s society. Urban consolidation centre (UCC) is a very popular solution for targeting these challenges (Allen et al., 2012; Katsela et al., 2022). UCCs are in this paper defined, in line with common descriptions, as terminals within a city or close to city borders that serve an entire city or specific areas or facilities (e.g. a mall) within a city (Allen et al., 2012), and that consolidates shipments from different shippers and carriers within the same vehicle (Benjelloun & Crainic, 2009). UCCs separate inbound, often long-haul, freight transport from short-haul deliveries within the city (Benjelloun & Crainic, 2009; Browne et al., 2005). This enables improved consolidation of goods, which can lead to higher fill rates on urban deliveries and a reduction in the number of freight companies operating. Furthermore, a change in transport mode to, for example, smaller and/or electric freight vehicles, which are more suitable for urban deliveries, is also a potential benefit when UCCs are employed (Quak & Nesterova, 2014; van Rooijen & Quak, 2010; Veličković et al., 2018).

Research on UCCs focuses to a large extent on their environmental and social benefits, thus addressing the negative impact on the natural environment in terms of e.g. outlets, and the impact on humans and the society at large. In a structured literature review of UCCs, Björklund and Johansson (2018) found that most of the studied articles argued for the environmental and social benefits as important driving factors. However, research focusing on the sustainability benefits of UCCs can be described as fragmented and transdisciplinary. There are strengths in addressing an area from multiple perspectives, however, this makes it difficult to compare the results from different studies. For example, the benefits of UCCs have been targeted in relation to other solutions that aim to improve urban distribution (Cherrett et al., 2017; Lebeau et al., 2018), decision tools that support the design of UCC solutions, such as modelling (Awasthi & Chauhan, 2012), and ways to estimate the environmental, social, and economic benefits of introducing UCCs by applying a cost – benefit analysis (Browne et al., 2007). Furthermore, the descriptions put forward of the environmental, social, and economic benefits of introducing UCCs can also be described as fragmented, applying different performance measurements, and presenting different results. For example, Cerutti et al. (2016) calculated the benefits to be a 30% reduction of CO₂eq, while Ambrosini et al. (2013) found only a 0.6% reduction in distance driven when implementing a UCC. Based on a review of the UCC literature, Lindkvist and Melander (2022) identified many different sustainability aspects targeted, such as improved quality of life, reduced number of vehicles, increased traffic safety, better working conditions, improved delivery services, and cost savings. As can be seen, the assessment of the benefits varies. Previous research has also highlighted some aspects that influence the potential benefits of implementing UCC solutions, such as location, government restrictions, vehicle types, and utilisation (Lin et al., 2016); as well as changes in fuel used (Björklund & Simm, 2019).

As the fragmented research targeting the sustainability benefits of UCCs continues to grow, the importance of providing an overview of the knowledge available increases. There is still a lack of awareness about the impact of these different aspects and which of them to consider in order to gain the most positive effects in the design of different UCC solutions. A systematic literature review (SLR) is a method for undertaking a thorough review of an area and to identify the dominant themes (Snyder, 2019; Tranfield et al., 2003). Such an approach is therefore relevant in this area because there exist clear gaps in the literature regarding the assessment of sustainability benefits and how the UCC setups affect these assessments. The need to provide explanations for why assessments of the benefits of implementing UCCs differ is central to guide future research and target the knowledge needs of practitioners. In response to this need, the purpose of this study is: To provide an overview of research on the sustainability assessment of UCCs, to describe dominant themes, identify underlaying factors why assessments differ and identify gaps for future research.

This paper is based on an SLR (see e.g. Denyer & Tranfield, 2009) of papers targeting the measurement of the sustainability benefits of UCCs. The motivation for selecting this method and the design of the review is presented in section 2. This is followed by a descriptive analysis (section 3) and a content analysis (section 4) of the identified articles. In order to compare the results of different studies, section 5 focuses on a cross-case analysis, of the results in different articles. The paper ends with a concluding discussion that proposes four factors influencing why the results of the various studies differ (section 6).

2. Methodology

Research attempting to estimate the sustainability benefits of UCCs are fragmented and transdisciplinary, and the benefits put forward point in different directions. Therefore, this paper takes its point of departure in a systematic literature review (SLR). An SLR contributes to the research area by providing a structured and comprehensible way of identifying relevant publications, summarising themes, and highlighting important gaps (Crowther & Cook, 2007; Denyer & Tranfield, 2009; Saenz & Koufteros, 2015; Tranfield et al., 2003). Identifying and analysing relevant publications can shed light on the issues that surround a diversified area, and can also direct future research.

One of the first steps in conducting an SLR is to define the unit of analysis (UoA) and to formulate the purpose and research questions, because these affect the search terms, inclusion criteria, and codes to be applied (Karjalainen & Juhola, 2021; Snyder, 2019). The UoA in this case consists of investigations of UCCs based on a measurement and assessment approach, and this is reflected in the search terms used. The search string was based on two search blocks, the first targeting UCCs and the second addressing measurements, because quantitative performance measurement is needed to capture the benefits of UCCs. In total, 20 different synonyms for UCCs were used (e.g. city/urban distribution cent*, city/urban terminal, and city/urban transhipment cent*), based on synonyms presented in Browne et al. (2005), Wolpert and Reuter (2012), and Björklund and Johansson (2018). The second block was targeted by applying synonyms such as assessmen* and quantif*, and quantitative measurements, such as CO₂, kilome*, and mile* to make the search more comprehensive. The database Scopus was used for the search and the terms were sought in the title, abstract, and keywords. Only peer-reviewed, published journal articles written in English were included, a common practice when conducting SLRs (Björklund & Johansson, 2018; Buldeo Rai & Dablanc, 2023; Gillström et al., 2024).

Figure 1 illustrates the process and in total, 85 journal articles were identified following the use of the search string. As a first step, the titles and abstracts were read by both authors individually, guided by the recommendation made by Tranfield et al. (2003), to exclude articles that do not fall under the scope of the study; in this case meaning that they did not target UCCs and their potentials. In general, there was a strong consensus between the authors, and if opinions differed, the authors discussed why that was so and if an article should be included or not until consensus was reached or. For the few instances the authors were in doubt regarding the inclusion or not, resulting in the inclusion of these articles in this phase. After this first exclusion phase, 51 articles remained. The next step was to read the full texts. During this read-through, four articles were excluded because they were outside the scope of this paper, which led to the final sample of 45 articles. Simultaneously, a coding manual was constructed (see e.g. Bryman & Bell, 2015) in Excel, to enable an analysis of the content and highlight similarities and differences. The first ten articles were read individually by the authors, with the purpose of identifying potential codes. These codes were then compared between the authors, discussed, and merged into one coding manual. This coding manual, together with the instructions to support the coding (Bryman & Bell, 2015), was also slightly adjusted during the reading of the rest of the articles in order to more completely encompass their content. The content analysis was based on the coding manual. Examples of codes are: input data applied (e.g. transport data); presenting a baseline with UCC or not; the type of UCC solution targeted (e.g. change of vehicle, location of UCC); the type of output (e.g. vkm, CO₂, NOx, external cost, traffic congestion); and if other measures were combined with UCC.

Figure 1. Illustration of the SLR process.

The analysis of the sample consisted of a descriptive analysis, a content analysis, and a cross-case analysis regarding the results reported in the articles. The descriptive analysis targeted aspects such as the journal, and the type of method used for data collection and analysis, which is common for SLRs (Björklund & Johansson, 2018; Denyer & Tranfield, 2009). The content analysis was based on the coding of the identified articles, where dominant themes were identified that were reoccurring throughout the sample (a common approach for SLRs, Snyder, 2019). For example, almost all articles included UCC set-ups or how UCCs have been assessed as well as variables applied, which points to this being dominant themes. Another example is that more than half of the papers target development of models and tools to make better assessments, making this a dominant theme.

The cross-case analysis of the calculations was conducted in order to facilitate comparison between the results of different studies, as many different performance indicators and UCC designs (e.g. change of transport mode or improved coordination) were applied. In order to perform the cross-case analysis, similar performance measurements are needed. For the analysis of this paper, the articles needed to present output data in vehicle kilometres (vkm) or a similar measurement that could easily be translated into vkm. The articles reported changes in vkm per day, per week, per month, or per year. To make the results comparable, they were recalculated to vkm per year under the following assumptions: 251 working days per year, 21 working days per month, and five working days per week (Workingdays.org, 2007). This analysis also provided the foundations for identifying factors that can explain why the assessments of UCC benefits differ. The process of identifying these factors can be described as pattern matching (see e.g. Miles et al., 2014), where the patterns consisted of articles that shared similar characteristics but reached different assessed results.

3. Descriptive analysis

The first article on UCCs was published in 2009 and the area has since then seen a steady increase in the number of articles, with approximately half of them published between 2019 and 2024. This development is in line with the increase in publications within the fields of transportation and city logistics/urban freight. The articles identified in this SLR were distributed across 36 different journals within several areas, verifying that the area is fragmented and transdisciplinary. However, more than half (26) of the articles were found within transport-oriented journals; for example: Transportation Research Parts A, C, and D (one article each), Transportation Research Part E (three articles) Research in Transport Economics (three articles), and Case Studies on Transport Policy (two articles). The remaining articles were scattered across several areas, such as journals focusing on the urban setting (e.g. International Journal of Urban Science, one article; and Sustainable Cities, one article), logistics and supply chain management (e.g. International Journal of Logistics Research and Application, one article), mathematical and modelling (e.g. Networks and Spatial Economics, two articles; and Mathematical and Computer Modelling, one article), or sustainability-oriented journals (e.g. Energies, two articles; and Sustainability, two articles).

The analysis also showed that, from a methodological perspective, there is less variety, with the majority focusing on secondary data and studies based on mathematical modelling. Several articles (21) are based solely on secondary data, and reports (e.g. road data, energy prices) are the most common source for data collection (20 articles). Transport data was acquired for ten articles, while primary data collection based on interviews or surveys was more seldom applied (five and seven articles, respectively). Regarding methods of analysis, mathematical modelling dominates (33 articles), followed by simpler forms of calculation (e.g. economic calculations, five articles). The remaining articles are conceptual or qualitative in nature. The large number of quantitative articles is not surprising due to the scope of the paper, even if it is worth noting the large number of articles that solely use secondary data.

Studies of cities in European countries dominate – Belgium (3), France (3), Italy (5), Poland, the Netherlands, Serbia (2), Slovakia, Spain (2), Sweden (2), the UK (5), and Ukraine – but some studies also focus on cities in Australia, China, India, Iran, Japan (2), Korea, and the Philippines. Only two articles include UCCs from different countries. Interestingly, a wide variety of sizes of cities have been studied, ranging from mega cities (e.g. Tokyo) and large (e.g. Rome), to medium and small cities (e.g. Novi Sad in Serbia and Malmö in Sweden).

4. Content analysis

A content analysis aims to identify and describe dominant themes in the literature. For this SLR, the themes have been divided into three sections: how articles contribute to improving models for optimising UCC solution design, different UCC set-ups and assessment, and type of performance measurements in focus in the different articles.

4.1. Improving models for optimising UCC solution design

Several articles aim to improve existing models used to simulate or optimise different forms of UCC, and thus do not provide additional insights into the sustainability benefits that can be achieved. These articles are an important basis for increasing the understanding of aspects that can be used to mathematically model the benefits of UCCs. Despite their aim to support the design of UCC systems, these articles do not necessarily provide any numbers on the input or output data provided. Instead, several articles aim to improve existing mathematical models by adding new aspects, or by providing input on how the data quality applied to different variables can be improved.

The location of a UCC is vital in saving distribution costs and minimising traffic congestion arising from the movement of goods in urban areas. This highlights the importance of applying high-quality mathematical models targeting the location problem. How to improve models targeting the location problem for UCCs is addressed in several articles, such as Cui and Xu (2022), which develops a method to address the lack of a standardised evaluation system for the UCC network location problem. The data presented in their article examines the location of existing UCCs, combined with the estimated logistics demand based on population density. Saeedi et al. (2019) propose a mathematical model (queuing theory-based bi-objective) that aims to optimise the environmental and economic costs of city logistics operations by making decisions about the UCC location problem. Awasthi et al. (2011) use fuzzy theory to quantify criteria values for location planning for UCCs; for example, logistics operators to deal with the uncertainty arising due to a lack of real data about location planning for new UCCs.

Several articles strive to develop more realistic input data to be applied in e.g. modelling. Lee and Chae (2021) target the distance and time for trips in order to model the vehicle routing problem. They take advantage of spatial data to analyse the shortest route possible between UCCs and receivers, and how this differs from reality when measuring closeness, and the differences in distance and time for outgoing and return trips. Hezarkhani et al. (2019) apply dispatch consolidation games to improve the mechanisms for managing cooperation, searching for ways to fairly allocate the obtained savings among the carriers using a UCC to consolidate their goods. Deng et al. (2021) also targeted how actors can share benefits using game-theoretical framework. Handoko et al. (2016) developed model to better include environmental sustainability using biobjective optimisation model. Finally, Lebeau et al. (2013) simulate the transfer from a conventional truck to an electric van, with a focus on the best combination of battery availability and loading capacity. Their results indicate that more traffic is generated by the van due to its limited payload. However, even though several interesting variables were included in their model, such as the number of vehicles, the resources needed to operate them and their availability, the frequency of trucks arriving at the UCC, and the quantity of freight being delivered and its nature (small volumes or large volumes), no concrete numbers were presented in the article.

4.2. Different UCC setups and assessment

Many of the studied articles use different UCC setups, which can affect the evaluation and assessment of their benefits. Increased consolidation is one of the main reasons for introducing UCCs and is included in all the articles to some degree. However, only a few studies estimate the effects in isolation. A UCC can also be an important facilitator for the change in transport modes, an aspect that is addressed in several studies. Adding to this are studies with a focus on how the benefits of using UCCs is affected depending on the location of them. A few articles investigate the benefits of UCCs in relation to other city logistics measures.

4.2.1. Consolidation

The consolidation effect is central and has often been the reason for implementing UCCs. Several articles explicitly investigated the effects of consolidation. Savchenko et al. (2022) apply mathematical modelling to evaluate the financial, social, and environmental costs of implementing micro-UCC terminals in the city of Kyiv in Ukraine. They found a potential reduction in social and environmental costs (in relation to aspects such as congestion, noise, and climate change) from 224 600 Euro without the use of the terminal to 126 100 Euro with its use. In a scenario study of six UCCs serving different parts of the city of Zaragoza in Spain, Escuín et al. (2012) found a decrease in transport distance from 400 km/day to 330 km/day (i.e. a 17% reduction). Kin et al. (2018) used mathematical modelling to calculate the (economic) costs of alternative distribution set-ups to find optimal solutions depending on how small and fragmented the volumes are, and the distances they need to be transported, and found UCCs to be the most cost-efficient solution when drop sizes are low and distances are long, while direct transport was more cost efficient for shorter distances.

4.2.2. Changes in the mode of transport

Since the UCC can be viewed as a decoupling point between inbound and outbound deliveries, it enables a shift in the mode of transport. Rail deliveries were one of the most targeted changes for inbound deliveries to the UCC (e.g. Kovač et al., 2023), and the change in outbound deliveries mainly targeted smaller distribution vehicles and electric vehicles. Roquel et al. (2018) assessed the benefits of the use of rail freight and a freight volume shift to use ports that are not located close to major cities (outer ports). However, they found the most effective solutions to be using ports while consolidating truck trips using UCCs at designated locations. This was because shifting transport to the ports reduces travel time, while consolidation improves freight operations, and reduces emissions. Pietrzak et al. (2021) applied mathematical calculations based on case-study data in a study focusing on the shift of urban deliveries from road to Light Freight Railway (i.e. electric trains) in the city of Szczecin, and found a potential decrease in external costs of more than 90%. Alessandrini et al. (2012) studied how UCCs can enable a shift to intermodal transport, i.e. combining road transport with rail, and based on road data from Rome in Italy, they found large potentials for improvement, with a 50% reduction in CO₂ emissions and a 52% reduction in energy used.

Other types of mode changes have also been studied. A pilot study in central London by Browne et al. (2011) included a mode change from diesel vans to electric cars and bikes that resulted in a decrease in distance travelled of approximately 20% (0.08 km per parcel) and a decrease in emissions of 80% measured in CO₂eq. Changing from vans to electric bikes for deliveries was also included in Ceccato (2023) who found that large savings of CO2 and externalities (up to 79%). With a focus on the production of electricity required for an electric delivery van to carry out its daily mission from a production plant located at the UCC, Napoli et al. (2021) applied a vehicle routing model (with time windows included in the problem constraints) to optimise freight distribution and thereby reduce the energy needed. In a scenario study on a limited area of Chicago, USA, Lin et al. (2016) investigated the benefits of applying vehicles of different sizes. A decrease in energy of 40% was identified as being the most positive scenario. However, an increase in energy was also found in scenarios where the baseline had a 100% fill rate and applied a larger vehicle. This study by Lin et al. (2016), who studied scenarios using different types of vehicles, provides an important lesson: the benefits vary hugely depending on the baseline (distribution before the implementation of a UCC). In cases with high fill rates, the implementation of a UCC can even generate negative effects in terms of energy needs. This is also in line with the findings of Browne et al. (2005), who showed that the benefits are highest in cases of low fill rate that apply direct transport. Lastly, Lemardelé et al. (2021) developed calculation models that could estimate the effects of having drones performing deliveries.

4.2.3. Changed location and number of UCCs

Where UCCs are located can have large effects on the delivery distances, which affects both the environmental and economic aspects. In the search for optimal locations for UCCs in Bratislava, the capital of Slovakia, Settey et al. (2021) applied mathematical modelling to investigate routes followed by electric and hybrid vehicles. However, although km for different routes and the number of transports is presented, no measurements are put forward to capture the potential change in total distance, the environmental impact, or energy use. In a scenario analysis, Veličković et al. (2018) found that the location and number of UCCs strongly affected the calculated external costs. For example, the authors found that only implementing one UCC in a larger city increased the external costs for all parameters used (when also factoring in the placement of the UCC). The largest decrease in external cost (24%) was found when the maximum number of UCCs were used (three in their study).

Some other articles on the location problem do, however, provide results clearly indicating more favourable locations and provide output data regarding the benefits that can be achieved. Li et al. (2018) applied a mixed-integer linear programming model to quantify and price the various negative externalities in order to determine the best location for UCCs based on data from Beijing on the location of 20 receivers (supermarkets) and the potential location of a UCC. They found that transportation costs account for 53% of the total, and balancing air pollution costs, environmental costs, and the capacity of the UCC, the optimal location for the UCC was found at the boundary between urban and suburban areas. Musolino et al. (2019) evaluated potential UCC locations by applying two planning levels. Firstly, potential locations are defined (public policy sets these on the basis of factors such as suitable land) and second the carriers’ behaviour is simulated (vehicle routing problem), based on variable restocking demands. Suitable locations taking into consideration cost and CO₂ outlets were identified. The authors reported a 14% difference in travel time when comparing two different locations.

4.2.4. UCCs combined with other city logistics measures

The inclusion of UCCs can also be combined with other types of measures, and in some cases these measures amplify the UCCs’ benefits. Friedrich and Elbert (2022) evaluated the impact of city toll schemes and different time windows in the use of a UCC located outside the city toll area. Their findings indicated that, even though tolls can increase the use of UCCs, there is likely to be a need to subsidize the UCC to present solutions that are attractive to the logistics industry. Furthermore, UCCs are less cost attractive for logistics service provider (LSPs) with larger delivery quantities per stop. Juvvala and Sarmah (2021) developed a mathematical model incorporating routing costs, vehicle fixed costs, emission costs, and zone entry fees in order to investigate the significant role of policy tools targeting purchase subsidy and zone entry fees in the promotion of electric vehicles in a UCC setting. Scenario 1 (purchase subsidy for EVs) was the only scenario indicating a decrease in transportation costs (cost/day) – of 5% – while all scenarios provided a reduction in CO₂, ranging from approximately 3% (scenario 3: vehicle taxes with an exemption for EVs) to 18% (scenario 2: zone entry fee with an exemption for EVs) to over 20% (Scenario 5: emission charges). This article provides no data on the changes in transport situation in the different scenarios.

Another combination is presented by Papoutsis et al. (2018), who investigated how retailers’ sustainability performance is affected and assessed in three urban freight solutions: a UCC, tethering (improved connectivity in the supply chain), and a Shared Bus. Their analysis of strengths and weaknesses showed that solutions requiring a significant initial investment and broad interventions are less sustainable, providing tethering with the highest sustainability score, e.g. an 8% reduction in CO₂. Taking a point of departure in the perspective of multiple actors, Lebeau et al. (2018) applied a multi-actor, multi-criteria analysis (MAMCA) to investigate various stakeholders’ perceptions of different UCC solutions. They found that the favourite solution for receivers, citizens, and the authorities was the use of electric vans, four UCCs, and night distribution with eco road taxes. But this was the worst alternative for the shippers. The authors reported a reduction of 2.9–8.1% in transport emissions, depending on the type of solution.

4.2.5. Integrated management

Another aspect examined in the content analysis is the benefits of integrated management in cities with multiple UCCs. Dupas et al. (2020) investigated a distribution network including carriers, huge shopping centres, and multiple UCCs in Tokyo. They found that a significant reduction in distance travelled (up to 39%) and emissions (up to 36% of CO₂ emissions) could be achieved by fully integrating the distribution network. Veličković et al. (2018) also studied the effects on external costs if an integrated planning system with multiple UCCs was used and found the largest benefits (more than 50% reduction in all external cost parameters), compared to when only a single UCC was used.

4.3. Performance measurements applied

The final theme of the content analysis concerns different performance measurements. Throughout the literature review, it was evident that the articles applied a wide variety of different performance measurements, in terms of both the sustainability benefits of the investigated measures and the data underlying the investigation. This is interesting to highlight because it makes the results from different articles difficult to compare. Different performance measurements were found in all three sustainability dimensions.

4.3.1. Changes in vehicle kilometres and load factors

Vehicle kilometres (vkm), as expected, is the most common measurement applied, and is very closely connected to environmental performance. The measurement was found in 19 of the studied articles and is used as both a result variable and an input variable in the calculations (e.g. Aljohani & Thompson, 2020). For example, Björklund and Simm (2019) described a 4% drop in vkm after the municipality started to use a UCC, and Paddeu (2021) found a reduction of 43 commercial vehicles each week, corresponding to a reduction of 52 vkm. On the other hand, driving distance was applied as an input variable by Cherrett et al. (2017), by using the shortest distance from the UCC for the consolidated distribution to the university studied.

Besides a reduction in kilometres driven, the load factor (i.e. fill rate) (see e.g. Anand et al., 2021; Dupas et al., 2023; Lebeau et al., 2018; Triantafyllou et al., 2014) and the number of vehicles applied (see e.g. Cherrett et al., 2017; Dupas et al., 2020) are other measurements that are frequently applied to capture the efficient use of vehicles.

4.3.2. Environmental and social aspects

Environmental impact is the most common result variable applied in the articles. Most commonly, the focus is on CO₂ or CO₂eq (e.g. Bukoye & Gadiraju, 2022; Musolino et al., 2019; Nocera & Cavallaro, 2017; Savchenko et al., 2022). But several articles also focus on other types of impact, such as NOx. For example, Paddeu (2021) found the highest social cost to be due to the emissions of NOx; and Katsela et al. (2022) found a large reduction in NOx when the UCC was combined with an additional micro terminal. Particles were also included as a measurement in several studies, such as Dupas et al. (2020) and Paddeu (2021). Furthermore, several of the articles integrate different environmental and social measurements in their attempts to present an overview expressed in total external costs (e.g. Estrada & Roca-Riu, 2017; Katsela et al., 2022; Papoutsis et al., 2018; Pietrzak et al., 2021). Katsela et al. (2022) included CO₂, NO_X, and PM, and Papoutsis et al. (2018) used the same measurements but also included noise and accidents (among other).

However, despite the widespread use of electric vehicles, energy use was not common among the performance measurements applied (see e.g. Katsela et al., 2022; Veličković et al., 2018). Exceptions can be found in, for example, Alessandrini et al. (2012), who estimated both the environmental (e.g. CO₂ and PM) and energy potentials associated with such a transfer. Pietrzak et al. (2021), on the other hand, applied a more qualitative approach that included more aspects, such as air quality, energy, service levels, profit, and road safety.

4.3.3. Economic aspects

Cost as a performance measurement has not been applied to the same extent as the social and environmental impacts of the implementation of UCCs (e.g. Anand et al., 2021; Escuín et al., 2012; Juvvala & Sarmah, 2021). One example is Hezarkhani et al. (2019), who provide a more comprehensive view by calculating the costs for the different actors involved. Interestingly, two articles (Janjevic & Ndiaye, 2017a, 2017b) only use cost as a result variable.

5. Results from the different studies

Twelve articles present sufficient data to make it possible to compare their results and calculation methods. Ten of these clearly present a baseline without a UCC, which makes it possible to include the changes that occurred when a UCC was implemented. In almost all cases, diesel trucks were applied in these baseline scenarios. Two additional articles presented data and calculations, but without a baseline. Descriptions of the twelve articles, including key characteristics of the system and type of deliveries, together with information about whether the vehicles in the UCC scenario differed from the baseline, can be found in Table 1.

Table 1. Description of the articles and changes made.

With a baseline	Description	Changed vehicle type
Dupas et al. (2023)	Scenario study of implementing UCC to handle deliveries in the city centre of Bordeaux	No
Katsela et al. (2022)	Three scenarios in the city of Malmö: (1) a micro UCC close to the delivery points, (2) consolidation at LSP terminals, and (3) both a UCC and a micro UCC.	Change to light electric trucks with reduced loading capacity for all scenarios.
Anand et al. (2021)	Simulation study of the distribution to retailers via one UCC in the city of Rotterdam.	Change to electric trucks with reduced loading capacity
Paddeu (2021)	Scenario study of deliveries to local business in Cagliari using a UCC.	Same electric trucks were used in both the baseline and scenario.
Roquel et al. (2018)	Scenario study of implementing 15 UCCs in the city of Manilla	Same vehicle used in both the baseline and scenario. No change in loading capacity.
Veličković et al. (2018)	Seven scenarios for three different locations of one to three UCCs in the city of Novi Sad. The two most extreme scenarios are: (1): one UCC at the “worst” available location, (2): three UCCs.	Change to smaller electric light distribution vehicles with reduced capacity.
Papoutsis et al. (2018)	Scenario study of one retailer’s deliveries to its stores in the City of Antwerp.	Change to slightly smaller trucks with reduced loading capacity.
Nocera and Cavallaro (2017)	Scenario study implementing a UCC and: (1) a few light electric trucks, (2) doubling the use of electric trucks, and (3) increased use of electric trucks by 67% compared to (2).	Change to light electric trucks, no change in loading capacity.
Alessandrini et al. (2012)	Scenario study of deliveries in Rome using a UCC in combination with inbound rail transport.	Same vehicle type for scenario 1. Change to hybrid vehicles in scenario 2. No change in loading capacity.
Browne et al. (2011)	A trial performed on deliveries from one office supply company to retailers in London.	Diesel-driven trucks were combined with electric trucks and electric tricycles, both with reduced loading capacity.
No baseline reported
Settey et al. (2021)	Scenario analysis on UCC locations based on one retailer’s deliveries to stores in Bratislava.	Same type of electric vehicles in all scenarios.
Dupas et al. (2020)	Scenario analysis of three carriers in Tokyo. Coordinated transport from LSP to the UCC, and integration (i.e. sending packages between UCCs) was tested. Two extreme scenarios are: Scenario 1: Integration, no pooling. Scenario 2: Pooling and integration. Study conducted in two parts, Part 1 with homogeneous vehicle fleet and part 2 with heterogeneous vehicle fleet.	Part 1: No change in vehicle or loading capacity Part 2: Vehicles with reduced loading capacity for all scenarios

The results from the different studies vary significantly, as outlined in Table 2. Here, additional calculations by the authors have been necessary, because the data in the articles was presented in different units, such as vkm/day, vkm/week, and vkm/month. The data was therefore recalculated into the unit vkm/year (using a database tracking the number of working days per year). Secondly, a calculation was made to present the benefits of each study as a percentage change regarding both environmental impact and vkm. The table is sorted from the study to present the most optimistic result first, down to the most pessimistic result from the articles (see Table 2).

Table 2. Comparison of the benefits of UCCs in different scenarios. A negative value means an increase compared to the baseline without UCC.

	Vkm/year Baseline	Vkm/year UCC	Change in vkm [%]	Change in CO2 [%]
Alessandrini et al. (2012), Scenario 2	72 037	27 610	62%	75%
Alessandrini et al. (2012) Scenario 1	72 037	27 610	62%	51%
Katsela et al. (2022) Scenario 3	43 700	17 100	61%	74%
Nocera and Cavallaro (2017) Scenario 3	3 600	1 400	61%	68%
Dupas et al. (2020) Part 1 Scenario 2	250 600	124 900	50%	48%
Settey et al. (2021) Scenario 1	397 700	203 900	49%	49%***
Settey et al. (2021) Scenario 2	397 700	232 000	42%	42%***
Dupas et al. (2023)	4 520 200	2 953 300	35%	25%
Nocera and Cavallaro (2017) Scenario 2	3 600	2 400	32%	33%
Dupas et al. (2020) Part 2 Scenario 2	316 900	217 100	31%	32%
Settey et al. (2021) Scenario 3	397 700	300 450	24%	24%***
Katsela et al. (2022) Scenario 2	43 700	34 8900	20%	29%
Browne et al. (2011)	35 000	28 500	19%	54%
Nocera and Cavallaro (2017) (2017) Scenario 1	3 600	3 100	14%	17%
Dupas et al. (2020) Part 1 Scenario 1	250 600	221 200	12%	5%
Dupas et al. (2020) Part 2 Scenario 2	316 900	285 400	10%	4%
Roquel et al. (2018)	829 600	795 000	4%	4%***
Anand et al. (2021)	12 300	11 900	4%	4%***
Papoutsis et al. (2018)	14 500	14 900	−3%	−7%
Veličković et al. (2018) Scenario 2	5 982 400	6 905 600	−15%	24%
Katsela et al. (2022) Scenario 1	43 700	62 700	−44%	−25%**
Veličković et al. (2018) Scenario 1	5 982 400	14 398 400	−141%	−98%**
Paddeu (2021)*	N/D	N/D	Reduction of 60 000 vkm/year

*The article has a baseline, but only reported a reduction in vkm.

**No reported CO₂ data, the number is based on the changes in external costs presented in the article.

***No reported CO₂ data, the number is based on the changes in vkm.

6. Identification of underlying differences between the studies

The reason for these large differences in vkm percentage changes between the various studies (as outlined in Table 2) can be traced to several factors, which are further elaborated below. Four factors are identified which make it difficult to compare results from the different articles and to find context independent assessment of the benefits of UCCs. Furthermore, the benefits are also strongly affected by the starting point, i.e. how well optimised the baseline is.

6.1. Changes in distribution vehicles

It is important to consider the changes in distribution vehicles when examining the benefits of implementing UCCs, especially if smaller vehicles and/or electric vehicles are used, as is common in the articles studied (see right-hand column in Table 1). This is evident in Veličković et al. (2018), who reported an increase of 15% in vkm in scenario 2; however, they found a reduction of 24% in external costs (the “best” scenario where three UCCs were used). This is due to electric distribution vehicles being introduced in the same calculations, which led to the decrease. Another example is found in Anand et al. (2021), who reported a reduction of 4% in vkm, but an increase in the number of vehicles needed, when changing to electric vehicles with reduced loading capacity. This is a case where the consolidation effect is so large that total vkm decreases, despite the increase in the number of vehicles. In a similar vein, Browne et al. (2011) found that, when diesel vehicles were abandoned in favour of electric vans and tricycles, the environmental impact could be reduced by 54%, and the total vkm was reduced by 19%. However, the outbound distance increased by 350% due to the smaller capacity of tricycles and electric vans. Switching to smaller and electric distribution vehicles has previously been highlighted in the literature (Quak et al., 2016), and our results indicate that the largest benfit is not the consolidation effect itself, but the opportunity to switch to smaller and electric distribution vehicles, which are more suitable for urban deliveries.

6.2. The scope of the system

The quantification of vkm clearly shows that the scope of the systems plays a large role when discussing the benefits of UCCs. One example of the different scopes applied is that some articles include both inbound and outbound transport, to and from the UCC, while others limit the study to only focus on the outbound transport. For example, Veličković et al. (2018) focused on only outbound transport from the UCC, while Papoutsis et al. (2018) studied inbound deliveries to the UCC from one retailer as well as the outbound transport to the recipient. Also, (Browne et al., 2011) included both outbound and inbound transport in their analysis, while others only focused on transport originating from the UCC (Katsela et al., 2022). Another way to change the scope of the system is to adjust the number of flows and receivers included, or the size of the city. For example, Settey et al. (2021) reported a reduction of almost 200 000 vkm/year, while Nocera and Cavallaro (2017), who studied the implementation of a UCC in a smaller city, only reported a reduction of 2 180 vkm/year for their scenario 3 (significantly increased use of electric trucks). Limiting their study to the transport of only one retailer, Papoutsis et al. (2018) found an increase of 455 vkm/year.

6.3. Number and location of UCCs

Another factor influencing the differences in the results of different studies is the number of UCCs included, and their locations. A larger number of UCCs does not, however, always guarantee larger benefit. For example, Roquel et al. (2018) studied scenarios with 15 UCCs, but only indicated a 4% reduction in vkm/year. The location of the UCC can also greatly influence the potential; for example, Settey et al. (2021) reported a 49% reduction in vkm when one UCC location was compared with another. Furthermore, Veličković et al. (2018) studied both the number of UCCs and different locations for them and reported vastly different results. All their scenarios showed an increase in vkm; however, the results ranged from an increase of 15% to an increase of 141%. It is evident that UCCs require resources in terms of, for example, cost and land use. As a consequence, if multiple UCCs are used, the amount of resources needed increases, and this is something that is usually not addressed in studies that target the number of UCCs.

6.4. Combined measures

Another factor that can affect the benefits is if UCCs are combined with other measures, such as using rail for the inbound transport, as in (Alessandrini et al., 2012), who reported a 51–75% reduction in vkm. Katsela et al. (2022) investigated scenarios including consolidation at LSPs’ terminals and the use of micro consolidation satellites and found that the benefits change in vkm ranged from an increase of 44% to a reduction of 62%. Another example is found in Nocera and Cavallaro (2017), who studied a scenario involving the introduction of a limited traffic zone, which forced the vehicles entering the city centre to be small and environmentally friendly. For this scenario, the authors reported a reduction of 61% in vkm and 68% in CO₂. Dupas et al. (2020) tested scenarios where the introduction of UCCs was combined with pooling (collaboration between LSPs) and integration (i.e. UCCs send parcels to one another), and found that the consolidation effects could be reinforced if additional solutions were included. The best scenario showed a 50% reduction in vkm.

7. Concluding discussion

Due to the fast-growing and dispersed field encompassing the benefits of UCCs, this article has attempted to provide an overview and identification of dominant themes through the use of a systematic literature review. The descriptive analysis confirms that it is a transdisciplinary area, with the identification of 45 articles published in as many as 36 different journals. By analysing the sample using a content analysis, three separate themes were identified, which show that the domains of UCC research target three areas: Improving models for optimising UCC solution design, different forms of UCC set-ups, and focusing on performance measurement.

Assessments of the benefits of UCC solutions varied greatly between the articles studied; for example, from a decrease of 62% vkm to an extreme of an increase of 141% vkm. However, most articles reported a reduction in vkm ranging from 4% to 62%. Even so, it is clear that the assessments of UCCs diverge significantly, and the differences can be traced to the following five factors:

• Different baselines, which could be more or less optimised. A very inefficient baseline will lead to larger benefits being identified than when a more efficient baseline is employed.

• Studies included changes in distribution vehicles in the UCC scenario, which often meant smaller vehicles with reduced loading capacity.

• The scope of the system differed; some articles included both inbound and outbound flows, while other only targeted the outbound flows from the UCC. Furthermore, some articles targeted deliveries to a whole city (and cities of different sizes), while others targeted deliveries from just one actor.

• The number of UCCs differed, as well as their location.

• UCCs were combined with other measures, such as the use of city tolls, which made it difficult to separate the effects of the UCC from those of the other measures, and presented the possibility that the measures might amplify each other’s benefits.

The results of this systematic literature review have several implications for research, especially since the review provides an overview of the area, which enables an understanding of what has already been done and what is still lacking. This study strives to bring clarity by highlighting several factors that affect the assessment of the benefits of using UCCs. Furthermore, the additional calculations made can facilitate comparison between studies and thus further increase the understanding of the field. The review also highlights that more than half of the papers target development of models and tools to make better assessments. This point to the complexity that surrounds the area. The second part of the implication for research constitutes identification of research gaps which can guide future research.

1. The first identified gap concerns factors that affect assessment of benefits. Several factors have been highlighted as having a significant impact on why the assessment of benefits for UCCs varies (as listed in the bullet points above). These factors can serve as guidance for future research, particularly emphasising the need for researchers to explicitly demonstrate how they account for these factors when conducting evaluations of UCCs. Additionally, future studies could delve deeper into understanding how the factors mentioned in the bullet list above specifically affect the benefits of implementing UCCs.

2. The second gap relates to a limitation that this study has identified, namely that the comparison of assessments in the articles is restricted to changes in CO₂ emissions and vkm since this was the most commonly available. Future research should continue investigating this limitation and consider incorporating other performance measurements beyond these two factors.

3. This research gap is associated with the bullet point concerning changes in distribution vehicles. While several studies have included electric distribution vehicles in their analyses and reported significant emissions reductions facilitated by the use of UCCs, the results are often intertwined with consolidation effects. This complexity makes it challenging to discern the most impactful factors. Future studies could address this by providing a clearer delineation and separately assessing the effects of distribution vehicle changes.

4. As the field continues to expand, there is increasing potential for making comparisons between different types of UCC solutions. For instance, the growing emphasis on micro-terminals and linked UCC micro-terminals (such as satellites), along with the rise in e-commerce and consumers as recipients, raises questions about whether the focus and calculation methods applied here differ from previous approaches. However, the limited number of articles available currently prevents a comprehensive comparison, leaving this as an area for future research.

5. This research gap concerns to the lack of transparency in the calculations made within the articles studied. This opacity significantly hinders the comparison of results, making the evaluation of UCCs difficult. Among the 45 studied articles, only twelve presented sufficient data to facilitate additional calculations for comparing different studies.

6. The sixth research gap emphasises the need for more empirical studies and real-life investigations (as opposed to relying solely on modelling or simulations). Conducting research in actual real-world environments allows us to capture the genuine benefits of UCCs.

7. This research gap concerns the lack of methodological diversity and the insufficient inclusion of primary data. Diversifying research methods and incorporating more first-hand data can enhance the understanding of UCCs’ impact.

Furthermore, there are also gaps within the field of UCC that have start to be bridged, such as the large amount of journal papers included in reference lists. This can be compared by the findings of Björklund and Johansson (2018) that stated that more than half of the references in their studied articles were reports or conference proceedings, instead of journal articles.

Implications for practice include highlighting the benefits for different types of UCC setups, even though the results of the review show that it is difficult to find a consensus. The review also contributes by showing factors that affect why the benefits can differ between different systems. This is important for policymakers and initiators of UCCs to understand and carefully contemplate in relation to UCC set-ups and their expectations. Furthermore, the largest benefits might not stem from consolidation effects, but rather from the opportunity to change vehicles. It is also evident that the benefits can differ depending on the stakeholder perspective in focus, which is addressed in some articles studied. This is something that policymakers need to account for.

Acknowledgement

This work was supported by the Swedish Energy Agency (project number: 2017-003184) and Swedish Electromobility Centre (project number: 12027).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Energimyndigheten and Swedish Electromobility Centre.