The LMA is legally defined as an association or a form of intermunicipality cooperation (Nunes Silva and Syrett 2006), comprising the nine municipalities of Lisbon district, north of the Tagus River (Amadora, Cascais, Lisbon, Loures, Mafra, Odivelas, Oeiras, Sintra, and Vila Franca de Xira), and the nine municipalities of Setúbal district, south of the Tagus River (Alcochete, Almada, Barreiro, Moita, Montijo, Palmela, Sesimbra, Setúbal, and Seixal). In figure S1 in the supporting information available on the Journal's website, a map of the LMA and the municipalities’ locations is provided.
Some of the municipalities are densely populated, such as Lisbon with 547,733 inhabitants in 2011, Sintra with a population of 377,835, Loures with a population of 205,054, and Amadora with a population of 175,136. Other municipalities are almost rural, such as Montijo with a population of 51,222 and Alcochete with a population of 17,569.
An MFA balance was made for the period 2003–2009 (on an annual basis) and the UMAn model was applied to produce values for materials accounting, throughput over time, distribution by economic activity, and spatial distribution. A comparison between the results for the LMA with the results from other urban regions was also performed.
Table 3 presents the DMC in the LMA, in 28 material categories. Dividing the materials into more than a handful of categories addresses gap 3 (limited discrimination of material types). Total DMC varies from approximately 22.5 million metric tons (tonnes) in 2003 to approximately 30 million tonnes in 2007, decreasing to 21.5 million tonnes in 2009. Approximately 45% of these materials were directly consumed (final goods), whereas the remaining had to undergo transformation before final consumption.
Table 3. Domestic material consumption (DMC) in the Lisbon Metropolitan Area by material type, 2003–2009 (tonnes)
|Low ash fuels||2,902,294||3,754,826||3,671,344||3,889,124||4,234,317||2,993,937||2,851,537|
|High ash fuels||63,608||80,611||76,450||64,472||82,092||68,769||60,028|
|Lubricants, oils, and solvents||759,929||1,026,142||960,439||779,512||864,702||632,509||589,720|
|Plastics and rubbers||353,487||426,245||411,457||344,036||417,780||465,708||402,413|
|Iron, steel alloying metals, and ferrous metals||942,265||1,164,525||1,147,964||967,148||1,222,615||1,128,539||1,089,759|
|Nonferrous heavy metals||116,550||144,938||140,323||118,077||134,255||126,160||113,337|
|Other (fibers, salt, or inorganic parts of animals)||111,487||131,290||115,083||99,745||108,683||82,148||76,236|
|Oils and fats||149,679||242,322||178,803||160,416||177,291||141,279||144,176|
|Wood and fuels||347,033||541,997||502,174||425,085||516,073||374,823||410,647|
|Paper and board||509,777||631,178||633,211||488,253||556,883||977,971||857,992|
|Chemicals and fertilizers||625,959||775,855||699,990||538,199||654,158||543,464||506,922|
|Chemicals and pharmaceuticals||167,038||201,765||207,061||163,036||208,092||193,884||173,807|
|Fertilizers and pesticides||443,585||552,837||471,181||356,334||422,052||329,754||314,208|
|Tonnes per capita||8.23||10.76||10.40||8.70||10.36||8.12||7.61|
Figure 2 shows the LMA material balance in 2005 and gives consumption, changes in stocks, and outputs to the environment. Nonmetallic minerals represent the main fraction of materials in both the overall input flows (50%) and the addition to the stock (more than 80%). Biomass and FFs are the second- and third-most relevant input materials. The second-most important material category added to stock is metallic minerals.
Figure 2. Lisbon metropolitan material balance in 2005 (million tonnes [Mt]). DMI = direct material input; DMO = direct material output; MSW = municipal solid waste; IW = industrial solid waste; C&DW = construction and demolition waste.
Download figure to PowerPoint
As described in the section titled “Calculator,” the model provides more detailed information than can be shown in figure 2. For instance, in the case of the LMA, the most important fraction of the domestic processed output is the carbon incorporated in emissions (50%), followed by dissipative flows (21%) and municipal (MSW) and industrial (IW) solid wastes (18%). Recycling represents only about 17% of the total solid wastes (MSW, IW, and construction wastes) produced in 2005.
For the input flows and increase in stock, most of the nonmetallic mineral material category is made up of stone, cement, and sand. The most important subcategories in each of the other major material categories—FFs, metallic minerals, and biomass—are as follows:
- Low ash fuels are the most important material in the FFs category (66%).
- Food, comprising agricultural and animal biomass, is the most relevant category of biomass consumption (67%), although wood and fuels, and paper and board, also have significant values.
- Iron and steel (75%) is the dominant metallic mineral category consumed.
In summary, the LMA has been consuming a large amount of construction minerals, although decreasing in the last couple of years. Additionally, nondurable goods, such as fuels and food, are consumed in large amounts, which has implications for immediate effects on atmospheric pollution and wastewater management. Less-significant amounts are associated with durable goods, particularly iron- and steel-based products that will have an effect on waste management in the future as a result of the accumulation in the material stocks in the LMA.
Overall, material consumption patterns were relatively stable from 2003 to 2009, without significant changes in the share of materials consumed, even though the total DMC varies from year to year. In 2008 and 2009, there was a slight gain in the relative weight of biomass consumed and a reduction in consumption of nonmetallic minerals. This is mainly the result of a reduction in the consumption of construction minerals. In fact, in 2008 and 2009, there was a significant reduction of finished new buildings along with an overall reduction in construction activities (INE 2012a).
The throughput dynamics was performed using the materials throughput matrix of the UMAn model, as described in the section titled “Throughput” (see figure 3). This analysis allows, for instance, prioritizing recovery goals for particular material types. According to the results, there are three main periods where three different material categories are more relevant in terms of potential amounts available for recovery:
- From 2010 to 2013: Products put into place between 2003 and 2009 that are FF based, such as oils and plastic, are retired and this material category comprises the bulk of the recyclable material flow for all products put in place during the 2003–2009 time frame.
- From 2015 to 2027: The retirement of products put into place between 2003 and 2009 that contain metallic minerals results in metallic minerals being the single largest contributor to the total recyclable material flow for products put into place during the 2003–2009 time frame, with two major peaks (2019 and 2024).
- From 2034 on: A steady increase in the retirement of products put into place between 2003 and 2009 that contain nonmetallic minerals is observed. This phenomenon expectedly extends beyond the studied time frame (2050), considering the average durability of products they are part of (e.g., buildings and infrastructures).
The dynamics shows a delay of approximately 40 years before the amount of recyclable materials from products put into place during the 2003–2009 time frame peaks. Nevertheless, even now, approximately 1 million tonnes of durable goods currently produced are potentially available for recovery every year. When compared to the amounts of wastes from durable goods recovered in 2005 (approximately 600,000 tonnes, according to the Urban Waste Statistics of INE for 2005; INE 2012b), approximately 400,000 tonnes of materials could still potentially be recovered in the region. These results illustrate how the UMAn model addresses gap 6 (the lack of understanding about the dynamics of added stock).
The next step is allocating the materials to different economic activities by creating the economic activity distribution matrix (equation (16)). The results given in table 4 demonstrate the ability of the model to address gap 4 (the limited resolution of consumption by economic activity). The economic activities of retail and wholesale trade consume the largest share of materials—approximately 15 million tonnes—followed by manufacturing activities, with more than 7 million tonnes (mostly FFs and nonmetallic minerals). The third-largest consumer of materials is construction, with approximately 3 million tonnes. The attribution of material consumption into different economic activities may provide valuable information to assess the resilience of the metropolitan area while providing information on the amounts of materials needed to supply the population and the regional economy. Additionally, if this allocation is coupled with information on the spatial distribution of consumption, allowing the identification of areas where specific materials are concentrated, it may unveil different dimensions of centrality within the LMA (see table 4). By incorporating the spatial distribution into the analysis (equation (19)), it can be seen that there is a large variation of the material consumption between the different municipalities in the LMA. There is a clear difference between the Lisbon district (the northern portion of the LMA), where the largest fraction of materials is consumed, whereas the Setúbal district (the southern portion) consumes very low material quantities, as would be expected because of the large difference in the populations of the two areas. According to the census data provided by INE in 2011 (INE 2013), the Lisbon district had 2.25 million inhabitants and the Setúbal district 0.85 million inhabitants.
Table 4. Domestic material consumption (DMC) per economic activity and municipality of the Lisbon Metropolitan Area, 2005
| ||DMC by statistical classification of economic activities (NACE) 1.1 category (tonnes)||DMC per person (tonnes per capita)||End Use DMC per person (tonnes per capita, weighted by relative purchasing power)|
|Municipality||A+B||C||D||E||F||G||H||I||J||K||L+M||N||O+P+Q|| || |
|Vila Franca de Xira||326||0||139,371||9,465||137,078||658,837||1,452||2,741||460||11,580||2,587||495||2,163||6.8||8.3|
|TOTAL||5,111||241,413||7,613,513||740,180||3,280,231||15,817,683||65,860||146,612||36,641||439,215||409,399||16,602||98,107|| || |
In per capita terms (see table 4), there is also a significant difference in DMC from one municipality to another, ranging from a minimum of 3.3 tonnes per capita in the Moita municipality to 26.2 tonnes per capita in Palmela. DMC in the municipality of Lisbon is 17.3 tonnes per capita. The difference between per capita values for the different municipalities is not as large when end-use DMC per capita values are compared; these values are weighted by the relative purchasing power of the inhabitants for each municipality in the context of the total purchasing power of the LMA. End-use DMC values range from 6.8 tonnes per capita in Moita to 16.3 tonnes in Lisbon.
Further, looking at the concentration of materials in the economic activities, three examples can be highlighted—Lisbon, Loures, and Palmela municipalities. In all three, the materials allocated to the economic activities greatly surpasses the material consumption allocated to the population, suggesting that these municipalities are potentially the main suppliers of materials within the LMA. Some facts may support this hypothesis, namely:
- The main regional wholesale market (MARL 2012) is located in Loures.
- Palmela, though an important agricultural area, is also important in terms of industry and infrastructure. The biggest automobile manufacturing cluster in the country exists in the region and important transportation infrastructure has been recently built, particularly the Lisbon-Palmela-Setúbal railway (Florentino and Nascimento 2009). This area will also be the future location of LOG Z, a third generation logistics platform (logz 2012).
- Lisbon is the capital city of Portugal, with many jobs and a large number of commuters (Niza et al. 2009), hence boosting the need for a higher concentration of economic activities to accommodate their consumption needs.
Though results from studies with different methods should be interpreted with caution, the numbers calculated for the LMA are within the range of values calculated for other European metropolitan areas, as shown in table 5. This table shows that per capita consumption estimates in the LMA are larger than the average of the table's studies for biomass and lower than average for metallic minerals. Also, these results indicate that the LMA is the lowest consumer of FFs and the highest consumer of biomass of the metropolitan areas in this table. The difference in FF consumption might be explained by the fact that stronger economies, such as Germany and the United Kingdom, usually tend to consume more FFs (e.g., to feed industries), whereas transitional economies, such as Portugal (Niza and Ferrão, 2006), tend to consume relatively more nonmetallic minerals (for construction) and biomass.
Table 5. Comparison of estimates of types of materials consumed in several metropolitan areas (tonnes/person)
| ||Material category|| |
| || || ||Minerals|| || |
|Lisbon Metro (DMC 2005)||2.12||1.84||0.52||5.39||0.53||10.40|
|York region (2000)a||1.01||3.26||0.15||7.51||0.01||11.94|
|Greater London (2000)b||1.65||2.05||0.13||3.91||0.73||8.47|
|Hamburg Metro (DMC 2001)c||2.05||4.29||6.00||−0.24d||12.10|
|Vienna Metro (DMC 2003)c||1.12||2.61||5.10||0.38||9.20|
|Average value for these studies||1.59||2.81||5.74||0.28||10.42|
In contrast to the other metropolitan area studies whose results are summarized in table 5, the UMAn model disaggregates the broad material categories so that more-specific estimates of the types of materials consumed is available, as shown in table 3. Further, none of the metropolitan area studies other than the UMAn LMA study provide an estimate of the changes in material stocks. Also, for the studies other than the UMAn LMA study, there is very little information about where in the metropolitan area the material flows occur (e.g., the Vienna, Hamburg, and Paris studies separated the study area into up to three concentric rings around the urban core). In addition, the studies other than the UMAn LMA study provided only highly aggregated estimates about the allocation of material consumption to economic activities (e.g., the York study assigns material flows to three categories of economic activity, whereas the UMAn model assigns material flows to more than a dozen categories of economic activity).
One test of the model's reliability is whether the estimates provided by the model result in a material balance for the region as a whole. For the LMA case study, the material input and output estimation is within 3.3% of closing the balance. In other words, the model's estimate of net addition to stocks accounts for nearly all of the difference between direct material input and direct material output. Further, when comparing estimations of FF produced by the model with direct data for sales of FFs from the Energy Directorate (DGEG 2013) the difference is approximately 1.2% for the overall period.