Illegal Use of Loading Bays and Its Impact on the Use of Public Space

: Loading bays are public spaces reserved for the operation of freight vehicles, and it is well known that there are signiﬁcant problems concerning their use due to non-compliance with existing regulations. Unlawful use of loading bays leads to double parking, or to parking on the pavement or in restricted areas. This article has two objectives: Firstly, the study and analysis of the use of loading bays (type of demand, parking duration, illegal use, etc.), as well as their use according to their morphology. Secondly, the quantitative assessment of the inﬂuence of illegal use with regard to the e ﬃ cient use of public urban space. Illegal use is quantitatively assessed by calculating the number of loading bays that are used inappropriately and the surface area (m 2 ) of public space used incorrectly. In the analysis carried out in the city of Santander (Spain), it can be observed that the urban morphology of loading zones inﬂuences their use: The greater the capacity of the loading zone, the less e ﬃ cient is its use. Moreover, it is observed that the degree of illegal use within loading zones is very high and that illegally excessive parking durations have a greater impact on the use of the ground space than vehicle type.


Introduction
The effects of urban freight transport on city mobility and congestion are significant, giving rise to such additional concerns as increases in air and noise pollution and accidents, among other things. Numerous research papers have attempted to minimize the negative impact this has, not only on traffic and congestion, but also on urban space consumption [1][2][3][4]. Thus, several measures have been proposed aimed at regulating freight vehicle access to cities [5,6], optimizing delivery routes in real time [7] through urban consolidation centers [8], regulating freight vehicle off-hour deliveries [9][10][11] and providing logistics infrastructure and facilities [12,13]. De Marco et al. [14] analyzed a dataset of 70 European cities that have been piloting or rolling out these and other measures, with the objective of providing an updated indication of the status of those measures. In this context, most urban areas reserve part of their public space to facilitate the loading and unloading of materials and/or goods, and these are called loading zones (L/U zones). Each loading zone is composed of a finite number of loading bays (LBs). In other words, the number of loading bays of one loading zone is equal to its capacity. The types of vehicle that may use loading zones in cities are determined by local regulations. The characteristics that such vehicles have may include vehicle weight, for example, in some Spanish cities [15], vehicle size or the levels of pollution emitted by the vehicle, as in some Italian cities [16].
Alho and e Silva [43] evaluated the service level of loading bays as well as the illegal parking rates in Lisbon and observed that non-freight vehicles accounted for 80% of demand.
As stated before, one of the consequences of the illegal use of loading bays is that, due to the lack of available space, freight vehicles opt to double park or to park on the pavement or in restricted areas, thereby using urban space inappropriately. Disturbances in city traffic flow and pedestrian mobility are a result of this incorrect use [37]. Alho et al. [44] performed a dynamic simulation with multiple configurations in loading zones, and achieved the restriction of loading zone use to authorized vehicles only to various degrees, so as to be able to compare the resulting disturbances to traffic flow.
Most of the above analyzed research focused on planning loading zone location, design or capacity and on analyzing the effects of illegal loading zone use on traffic. However, as far as the authors know, there is a lack of studies analyzing its effect on the use of public space (efficient use of loading zones implies an efficient use of public space) and how the different compliance rates to usage regulations may affect this efficiency.
For this reason, this article has two objectives: First, to analyze loading zone use in an urban context through the exhaustive analysis of representative loading zones based on the Santander (Spain) city center. The indicators analyzed include: Arrival time, parking duration, vehicle type and parking location, among others. The use of these loading zones is thereby studied according to their morphology in order to understand the degree and type of illegal use taking place, among other data. Based on this analysis, the second objective of this article is to quantitatively assess the influence of such illegal actions on the use of urban public space. To this end, the number of loading bays and the surface area of incorrectly used urban public space (in m 2 ) is obtained for multiple scenarios (different degrees and types of illegal use).
The introduction and state of the art have been presented in this section. Section 2 describes the methodology developed. The data collected and the analysis and comprehension of loading zone usage is carried out in Section 3. With this analysis, the quantitative assessment of the influence of illegal use is developed in Section 4. Finally, the main conclusions are set out in Section 5.

Methodology and Case Study Description
The methodology proposed is based on an exhaustive analysis of the data collected on the vehicles using the studied loading zones. Both the characteristics of the loading zones themselves and their use, meaning who uses them and how they are used, are analyzed, differentiating between their legal and illegal use. A discrete event modeling approach is then applied to simulate differing private user and freight vehicle compliance. These data are later applied in order to quantitatively assess the consequences of incorrect loading zone use, both legal and illegal, and to determine the degree of influence exerted by each non-compliance with existing regulations ( Figure 1).
As loading zone use may vary not only across cities and/or countries, but also within a city according to the loading zone type, quantity, location, etc., a representative sample of the use of loading zones in the area to be analyzed is essential. Firstly, it should be verified that both freight and passenger vehicles have parking problems in the area. Secondly, data from a representative sample of loading zones should be collected. Studied areas should have different characteristics, both within the loading zone itself and concerning the street where it is located.
The loading zones of the Santander city center were analyzed following this approach. This area, like most urban centers, has significant parking problems [45]. The main land use is commercial and retail use, as well as areas used by restaurants, hotels and offices, and, to a lesser extent, residential and educational use. The Santander city center has both pedestrian and traffic streets. There are 48 loading zones, all of them are curb-side parking spots, i.e., the parking is parallel to the pavement. The 48 loading zones were grouped according the following attributes: • Location and size of establishments obtained from the Santander Town Hall. • Location, capacity and regulation of the loading zones, obtained from the Santander City Council and field data.

•
Existence and characteristics of on-street parking, pavement and physical obstacles.

•
Characteristics of the Santander transportation network, e.g., types of streets and intersections, provided by the Santander City Council and field data.
Sustainability 2020, 12, x FOR PEER REVIEW 4 of 17  Characteristics of the Santander transportation network, e.g., types of streets and intersections, provided by the Santander City Council and field data. Four groups were built according to the attributes described above. Then, 652 surveys were done on the loading zones of the center of Santander. The proposal was to collect general data about the uses of the loading zones and their surroundings, not only from the surveys given to carriers, but also from the observations collected while the surveys were done. The general collected data confirmed that the loading zones were well classified by the four groups. Furthermore, it indicated what the representative loading zone of each group was based on the behavior of the carriers and the vehicles forbidden in the loading zones (e.g., cars).
Therefore, four loading zones underwent in-depth analysis within this study because they had different characteristics encompassing all of the variety existing in the city center. Information on these loading zones, which covers 19 loading bays, is summarized in Table 1. The locations of the analyzed loading zones are shown in Figure 2. Loading zones are identified throughout the article using the name of the street where they are found. Furthermore, their capacity is stated as "capacity", understood as the number of loading bays in each loading zone. Four groups were built according to the attributes described above. Then, 652 surveys were done on the loading zones of the center of Santander. The proposal was to collect general data about the uses of the loading zones and their surroundings, not only from the surveys given to carriers, but also from the observations collected while the surveys were done. The general collected data confirmed that the loading zones were well classified by the four groups. Furthermore, it indicated what the representative loading zone of each group was based on the behavior of the carriers and the vehicles forbidden in the loading zones (e.g., cars).
Therefore, four loading zones underwent in-depth analysis within this study because they had different characteristics encompassing all of the variety existing in the city center. Information on these loading zones, which covers 19 loading bays, is summarized in Table 1. The locations of the analyzed loading zones are shown in Figure 2. Loading zones are identified throughout the article using the name of the street where they are found. Furthermore, their capacity is stated as "capacity", understood as the number of loading bays in each loading zone.  The characteristics of both the loading zones and the streets where they are located is shown in the following table:  The characteristics of both the loading zones and the streets where they are located is shown in the following table: The use of loading bays is regulated by the Traffic Ordinances issued by the Santander City Council [15]. These ordinances establish the following regulations in the four loading zones studied:

•
Type of freight vehicles permitted: Vehicles with a gross vehicle weight rating (GVWR) between 1.8-8 tons. Finally, the data collection was carried out by: • Recordings from seven business days (from Monday to Saturday) between 7:00 a.m.-2:00 p.m. in the four loading zones studied. They were undertaken with the aim of knowing how vehicles that use the loading bays behave, their types and usage habits.

•
Six hundred and fifty-two surveys during a single day in 10 different loading zones, which covered 56 loading bays, in the center of Santander. These surveys were used to verify that the sample of the four loading zones was indeed a representative sample of the city center.

Vehicle Type and Arrival Rate
The records collected in both loading zones provided information on the type of vehicles that used these loading bays. Vehicles were classified by their characteristics exactly as shown in the following table (Table 2). It was observed that the freight vehicles whose characteristics did not match the permitted span did not use the loading bays. No illegal use of this kind was observed. Conversely, cars did use the loading bays despite the fact that this was prohibited ( Figure 3). It was observed that a large number of cars used the loading bays, at a rate reaching 49%, in the Jesus de Monasterio loading zone. This means that the number of available spaces for freight vehicles was reduced.
Sustainability 2020, 12, x FOR PEER REVIEW 6 of 17  Recordings from seven business days (from Monday to Saturday) between 7:00 a.m.-2:00 p.m. in the four loading zones studied. They were undertaken with the aim of knowing how vehicles that use the loading bays behave, their types and usage habits.  Six hundred and fifty-two surveys during a single day in 10 different loading zones, which covered 56 loading bays, in the center of Santander. These surveys were used to verify that the sample of the four loading zones was indeed a representative sample of the city center.

Vehicle Type and Arrival Rate
The records collected in both loading zones provided information on the type of vehicles that used these loading bays. Vehicles were classified by their characteristics exactly as shown in the following table (Table 2). It was observed that the freight vehicles whose characteristics did not match the permitted span did not use the loading bays. No illegal use of this kind was observed. Conversely, cars did use the loading bays despite the fact that this was prohibited ( Figure 3). It was observed that a large number of cars used the loading bays, at a rate reaching 49%, in the Jesus de Monasterio loading zone. This means that the number of available spaces for freight vehicles was reduced. Regarding the arrival rate, it was therefore analyzed during the time period when loading bays were reserved for freight vehicles (7:00 a.m.-2:00 p.m.). The arrival of freight vehicles is shown in Figure 4, calculated as an average of seven business days. As can be seen, the arrival of freight vehicles was slightly reduced throughout the four loading zones during the earliest hours. In the La Leña Square and Cervantes Street loading zones, demand was more or less constant throughout the morning. The Jesus de Monasterio and Vargas loading zones, in contrast, demonstrated an increase in demand between 11:00 a.m. and 12:00 p.m. Regarding the arrival rate, it was therefore analyzed during the time period when loading bays were reserved for freight vehicles (7:00 a.m.-2:00 p.m.). The arrival of freight vehicles is shown in Figure 4, calculated as an average of seven business days. As can be seen, the arrival of freight vehicles was slightly reduced throughout the four loading zones during the earliest hours. In the La Leña Square and Cervantes Street loading zones, demand was more or less constant throughout the morning. The Jesus de Monasterio and Vargas loading zones, in contrast, demonstrated an increase in demand between 11:00 a.m. and 12:00 p.m.

Parking Duration and Loading Zone Occupation
The ordinance that regulates the loading bays analyzed states that vehicles may not be parked for more than 30 min. When the percentage of vehicles that violate this regulation on parking time

Parking Duration and Loading Zone Occupation
The ordinance that regulates the loading bays analyzed states that vehicles may not be parked for more than 30 min. When the percentage of vehicles that violate this regulation on parking time limits was analyzed, we observed that nearly 30% of the vehicles in Cervantes Street exceeded the 30-min limit: 16% were parked for more than 60 min, equivalent to more than double the duration allowed. In consequence, vehicle rotation in the loading bays decreased. The reasons why the time limits were exceeded are unknown, however, it was observed that less than 5% of the vehicles that were parked for more than 30 min were due to loading and unloading activities.
Parking duration is shown by vehicle type in Figure 5. The average of the four loading zones analyzed is shown because the differences among them were minimal. The probability that light freight vehicles (LFs) exceeded the permitted time limit was approximately 30%. The probability of exceeding 60 min was approximately 15%.

Parking Duration and Loading Zone Occupation
The ordinance that regulates the loading bays analyzed states that vehicles may not be parked for more than 30 min. When the percentage of vehicles that violate this regulation on parking time limits was analyzed, we observed that nearly 30% of the vehicles in Cervantes Street exceeded the 30-min limit: 16% were parked for more than 60 min, equivalent to more than double the duration allowed. In consequence, vehicle rotation in the loading bays decreased. The reasons why the time limits were exceeded are unknown, however, it was observed that less than 5% of the vehicles that were parked for more than 30 min were due to loading and unloading activities.
Parking duration is shown by vehicle type in Figure 5. The average of the four loading zones analyzed is shown because the differences among them were minimal. The probability that light freight vehicles (LFs) exceeded the permitted time limit was approximately 30%. The probability of exceeding 60 min was approximately 15%. The occupation of loading zones and their surrounding areas (double parking, parking in restricted areas and on the pavement) throughout the morning (7:00-14:00) is shown in Figure 6. The average occupation of each zone can be seen in the stacked line chart (average of the seven days analyzed). The bars display maximum occupation. The occupation in L/U zones could be by any type of vehicle (freight vehicle or cars). Both the maximum and average occupation in the L/U zones are shown in blue. The surrounding areas could be also be illegally occupied by any type of vehicle, and in this case, this is represented in green. Furthermore, the capacity of the L/U zones is represented with a yellow line for quick comparison with the current occupation.
The average occupation of each loading zone, in terms of capacity, is observed in the graphs found in Figure 6. In other words: The occupation of loading zones and their surrounding areas (double parking, parking in restricted areas and on the pavement) throughout the morning (7:00-14:00) is shown in Figure 6. The average occupation of each zone can be seen in the stacked line chart (average of the seven days analyzed). The bars display maximum occupation. The occupation in L/U zones could be by any type of vehicle (freight vehicle or cars). Both the maximum and average occupation in the L/U zones are shown in blue. The surrounding areas could be also be illegally occupied by any type of vehicle, and in this case, this is represented in green. Furthermore, the capacity of the L/U zones is represented with a yellow line for quick comparison with the current occupation.
The average occupation of each loading zone, in terms of capacity, is observed in the graphs found in Figure 6. In other words:

•
Average loading zone occupation exceeded capacity in the case of Jesus de Monasterio Street, where the maximum occupation during the better part of the morning was one vehicle more than capacity.

•
Average loading zone occupation never reached capacity in the case of Vargas Street, where the maximum was one vehicle less.

•
Average loading zone occupation was slightly less than capacity in Cervantes Street.

•
Average loading zone occupation was approximately equal to capacity in the case of La Leña Square.
Based on the data collected, the following issues were observed to contribute to the inefficient use of loading zones: • The greater the zone capacity, the less efficient its use is. This is due to the fact that freight vehicles do not always park in an organized way to optimize zone use. It was observed that when capacity was greater than five vehicles, the loading zone rarely reached capacity.

•
The zone morphology, that is, the type of obstacles on either side of the loading zone, if applicable, influences its use. In sum, the most optimal use of the loading zone is observed when there are no   As has been demonstrated above, the current use made of loading bays did not comply with the ordinance that regulates them and the number of infractions committed in loading bays was high. Infractions in loading bays occur due to the types of vehicle that use them and/or the time that said vehicles are parked in them. As a result of the inappropriate use of loading bays, freight vehicles do not have the space they need in order to carry out their activities. Other illegal uses of space are chosen to this end, such as double parking or parking on pavements or in restricted areas. The percentage of vehicles that used loading bays and their surroundings illegally during the study period (7:00 a.m.-2:00 p.m.) is shown in the following table (Table 3) by illegal use.

Scenario Definition and Simulation
The data collected in loading zones confirms the expected: loading bays are illegally used to a large degree, with illegal use observed as follows: • Use of loading bays by unauthorized vehicles. • Non-compliance with the maximum time limits.
As a result of these two unlawful actions, vehicles double park, park on the pavement or park in restricted areas, thereby committing another illegal action. In order to quantitatively assess the consequences of the inappropriate use of loading bays, the following scenarios were analyzed, comprising various degrees of compliance and types of illegal use (Figure 7 Scenarios D, E and F represent those scenarios in which loading bays are monitored but the application of regulations is not strict. The consequences of the illegal use of loading bays was quantitatively assessed for these seven scenarios through the study of their effect on the use of public space. This study was performed through the calculation of the number of loading bays used inappropriately at peak times.
All the scenarios were applied to the four loading zones analyzed. This analysis was carried out based on the simulation of the movement of vehicles in loading zones through discrete event simulation; these events were defined as the presence of freight vehicles in loading bays. Rockwell Arena software, a discrete event simulator, was used. This software has already been applied in a wide range of research, including traffic simulation [44]. It allows for simulating various scenarios with different stochastic variables, such as the loading time by merchandise type and the arrival time distribution. The simulation process went as follows: Step 1: Vehicles are characterized by type.
Step 2: Vehicles arrive at the loading zone according to an arrival distribution.
Step 3: If space is available, they carry out their operations for a specified duration of time. Loading times vary according to the vehicle type and scenario being analyzed.
Step 4: If the loading zone is not available, vehicles will double park or park on pavements or in restricted areas.  Scenarios D, E and F represent those scenarios in which loading bays are monitored but the application of regulations is not strict. The consequences of the illegal use of loading bays was quantitatively assessed for these seven scenarios through the study of their effect on the use of public space. This study was performed through the calculation of the number of loading bays used inappropriately at peak times.
All the scenarios were applied to the four loading zones analyzed. This analysis was carried out based on the simulation of the movement of vehicles in loading zones through discrete event simulation; these events were defined as the presence of freight vehicles in loading bays. Rockwell Arena software, a discrete event simulator, was used. This software has already been applied in a wide range of research, including traffic simulation [44]. It allows for simulating various scenarios with different stochastic variables, such as the loading time by merchandise type and the arrival time distribution. The simulation process went as follows: Step 1: Vehicles are characterized by type.
Step 2: Vehicles arrive at the loading zone according to an arrival distribution.
Step 3: If space is available, they carry out their operations for a specified duration of time. Loading times vary according to the vehicle type and scenario being analyzed.
Step 4: If the loading zone is not available, vehicles will double park or park on pavements or in restricted areas.
The seven scenarios discussed were simulated by applying this methodology. First, the model was calibrated to replicate the current situation; the validation step was based on a reasonable adjustment test (Pearson's chi-squared test), confirming that the results obtained in the simulation were adapted to the values obtained with a confidence interval of more than 95%. Then, the other scenarios were simulated based on the current situation with their respective schemas. Performance indicators wwere obtained from the simulations, such as average loading zone occupation.
Loading zone capacity, or the number of loading bays needed for each of the scenarios, was obtained from the results of the simulations. The number of loading bays obtained from the The seven scenarios discussed were simulated by applying this methodology. First, the model was calibrated to replicate the current situation; the validation step was based on a reasonable adjustment test (Pearson's chi-squared test), confirming that the results obtained in the simulation were adapted to the values obtained with a confidence interval of more than 95%. Then, the other scenarios were simulated based on the current situation with their respective schemas. Performance indicators were obtained from the simulations, such as average loading zone occupation.
Loading zone capacity, or the number of loading bays needed for each of the scenarios, was obtained from the results of the simulations. The number of loading bays obtained from the simulation was compared with the current number of loading bays. The difference between them gave the number of vehicles that were using public space incorrectly. These vehicles could be found in the loading bays themselves or in adjacent lanes (vehicles that were double parked) or parked on pavements or in restricted areas. Using this information, as well as the average space that vehicles occupied, we were able to estimate the space that was being used incorrectly due to the varied illegal uses taking place in loading bays.

Scenarios Assesment
This methodology was applied to four loading zones that were analyzed for the seven study scenarios. The data collected and analyzed in Section 3.1 were reused and analyzed for this purpose, these being vehicle type, demand, occupation, parking duration by vehicle type, etc.
The data collected in the loading zones were gathered and analyzed, and the methodology was applied, obtaining the following indicators: -Average loading zone occupation for each of the seven scenarios. -Average number of vehicles illegally parked (double parked (DP), parked in a restricted area (RA), parked on the pavement (OP)) for each of the seven scenarios, taking into account that the number of loading bays is equal to the current number.
-Number of loading bays necessary for there to be no vehicles parked illegally (DP, RA, OP). -Number of loading bays being used inappropriately or not used efficiently due to illegal actions.
In order to simulate scenarios C and D, in which vehicles do not park for longer than 30 min, vehicles that exceeded this limit were hypothesized as parked for between 25 and 30 min. The seven aforementioned scenarios were simulated according to these conditions. One hundred (100) replications were run in each simulation scenario in order to account for random scenario factors. In the proposed model, the CPU time on an Intel Core i5 processor is about 30 s for each scenario. In the following graphs, (Figure 8) the average occupation of each loading zone and its surroundings is presented for the seven scenarios and the four loading zones.
The data collected in the loading zones were gathered and analyzed, and the methodology was applied, obtaining the following indicators: -Average loading zone occupation for each of the seven scenarios. -Average number of vehicles illegally parked (double parked (DP), parked in a restricted area (RA), parked on the pavement (OP)) for each of the seven scenarios, taking into account that the number of loading bays is equal to the current number. -Number of loading bays necessary for there to be no vehicles parked illegally (DP, RA, OP). -Number of loading bays being used inappropriately or not used efficiently due to illegal actions.
In order to simulate scenarios C and D, in which vehicles do not park for longer than 30 min, vehicles that exceeded this limit were hypothesized as parked for between 25 and 30 min. The seven aforementioned scenarios were simulated according to these conditions. One hundred (100) replications were run in each simulation scenario in order to account for random scenario factors. In the proposed model, the CPU time on an Intel Core i5 processor is about 30 s for each scenario. In the following graphs, (Figure 8) the average occupation of each loading zone and its surroundings is presented for the seven scenarios and the four loading zones. As observed in Figure 8, in general, the same excessive parking duration existed in scenario B as in the current situation even though it did not allow cars to use the loading bay; this scenario hardly improved the current situation. The only loading zone in which a significant improvement occurred As observed in Figure 8, in general, the same excessive parking duration existed in scenario B as in the current situation even though it did not allow cars to use the loading bay; this scenario hardly improved the current situation. The only loading zone in which a significant improvement occurred was in the Jesus de Monasterio loading zone. This is owed to the high percentage of illegal use in connection with vehicle type (cars), compared to illegal use in terms of excessive parking duration, as well as compared to the other three zones, as described in Table 3. This is why it can be deduced that the effect of illegal use by vehicles due to excessive duration is greater than the effect of illegal use due to vehicle type, taking into account current parking durations ( Figure 5).
If we observe scenarios D, E, and F, we see that the degree of illegal use was close to zero. Differences among the three scenarios were minimal, similar to the difference compared to the ideal scenario, scenario G, in which compliance with existing regulations was strictly observed. The Jesus de Monasterio loading zone continued to be the exception for the same reason as mentioned above: the high degree of illegal use due to vehicle type, which was triple the amount of illegal use due to excessive parking duration.
Once the average loading zone occupation in the seven scenarios was obtained, we could derive the average loading bays that were being used inappropriately during peak times in each loading zone, as well as the 95% confidence interval half-width (Table 4). A loading bay has to be used whole, in other words, a freight vehicle needs one loading bay and less of one is not useful for it because it does not have space to park. For this reason, the rounded-up number of loading bays that were being used inappropriately and an estimation of the surface area of public space (in m 2 ) are also presented. The surface area measurement in m 2 wasestimated based on the dimensions of the vehicles discussed above (Table 2) and the vehicle types that use each zone (Figure 3), generating an average vehicle occupation of 9.6 m 2 . Inappropriately used space is understood as the space that is not being used efficiently or that is used for a purpose for which it was not designed, whether it is a space in a loading zone or one where double parking, parking on the pavement or parking in a restricted area take place. When the results obtained were analyzed, we observed that, in the current situation, an estimated total of 76.8 m 2 were incorrectly used throughout the total loading zones analyzed and their surroundings. Moreover, even in an ideal scenario, space within the Vargas loading zone will always be used inappropriately for the reasons discussed previously: This area never achieves capacity (eight freight vehicles) because the vehicles are not able to organize themselves and optimize the use of space, and therefore, there is always unused space. Therefore, this space could be available for other purposes, such as parking spaces for bikes, cars or motorbikes, or for underground rubbish bins, wider sidewalks or additional street furniture.
Furthermore, the percentage of illegalities, by duration and type, were similar-except in the case of Jesus de Monasterio Street, in which the duration was less than type illegalities (Table 3). It was observed that illegalities by parking duration (scenario B) had an equal or a greater impact than illegalities by type of vehicle (scenario C).
These results can be generalized to the whole city center, where there is very little public space, in order to estimate the public space that is currently being used inappropriately. The city center has 48 loading zones, which have been classified into five groups: Four of these were grouped due to their similarities with the four loading zones analyzed in terms of the type of illegal use that would occur there if the loading zone was full. In the fifth group are loading zones in which illegal use of this kind never occurs due to the characteristics of the street. Furthermore, it is known that 10 loading zones have a capacity greater than five. Based on the data analyzed and the results obtained, it is believed that vehicles are unable to organize in loading bays with a capacity of more than five bays, and one of the bays is not used properly. The number of public spaces that are used inappropriately and the average m 2 of inappropriately used public space throughout the city center can thereby be estimated. The space used incorrectly in the surroundings of loading zones is specified (Table 5). From this analysis, it can be estimated that a total of 622 m 2 of the Santander city center is currently being used incorrectly or inadequately due to loading bay use only. Moreover, even if there are no illegalities, the area of ten loading bays would be unused-96m 2 -one loading bay unused in each loading zone with a capacity greater than 5 m 2 . Furthermore, it has been seen that, as a consequence of the characteristics of the surroundings of the loading zones, inappropriate use varies in each zone. The average road that is inappropriately used by the vehicles that double park reaches 125 m 2 in the peak times, which may cause traffic disruption. There is more inappropriately used space that impacts on traffic than on pedestrians. The average pavement area that is incorrectly used is only 38 m 2 , nevertheless, is important for pedestrians. In addition, restricted area that is incorrectly used can generate problems for other users, for example, for buses or taxis, if the restricted area has bus or taxi stops, or for private vehicles if it is an exit from a garage.
Finally, knowing that the city center loading zones comprise 2121 m 2 , approximately 30% of the urban public space reserved exclusively for the loading and unloading of goods is being used incorrectly. It is possible that the incorrect use is not directly in the loading and unloading zones, but also in surroundings, but it is also an incorrect use.

Conclusions
The analysis carried out in this article has highlighted the illegal use of loading bays in an urban context through the analysis of the use and characteristics of loading zones of representative loading zones in the city of Santander (Spain). Using the information obtained, different regulation compliance scenarios have been simulated to assess the consequences of illegal loading bay use and its implication on efficient urban space use.
From the analysis of the use of the four loading zones, it has been observed that the degree of illegal use is very high: between 58% and 44% of vehicles make some kind of unlawful use of the loading bay, whether due to type and/or excessive time durations. Hence, between 34% and 55% of vehicles commit other illegal acts in the spaces adjacent to loading zones, such as double parking or parking on the pavement or in restricted areas. On the other hand, with respect to the quantitative assessment of the inappropriate use of loading bays, it was observed that illegal use due to excessive duration has a greater impact than illegal use due to vehicle type (provided that the percentage of illegal use is similar).
Additionally, the following conclusions have been reached, which may help to define loading zone capacity and characteristics: • The larger the loading zone capacity, the less efficient its use is. This is due to the fact that the bigger the capacity, the less probable it is that vehicles will park there in an organized way.

•
With regard to loading zone characteristics, it was observed that the characteristics of the space on either side of the loading zone influence the use of the zone itself.
The conclusions obtained in this investigation provide several recommendations for the city authorities. These recommendations or strategies for the design of loading zones are:

•
The capacity of the loading zone has to be five at most, in other words, five medium freight vehicles parked at the same time.

•
Avoid loading zones without pavements or physical objects. If it is necessary to place a physical object on either side of a loading zone in order to protect street furniture, such is waste bins, authorities should use a concrete slab instead of bollards or any other item.

•
Authorities should implement measures to control loading zones. Not only to control the type of vehicles that use it, but also to control the duration of parking in order to promote vehicle rotation.

•
It is advisable for regulations to be strict so that inappropriate use of urban public space is reduced, both in and around loading zones. Funding: This study and the development of future research are possible thanks to financing from the Santander City Council from the MOBIDATA-SDR project.

Conflicts of Interest:
The authors declare no conflict of interest.