REFERENCES - The ADVANCE project | Advance



Modelling Transport Infrastructure in IAMsThis methodology is written for modelers who want to implement transport infrastructure in IAMs. During general development of the IMACLIM-R model over the past number of years it has been decided that the physical transport infrastructure e.g. roads should also be incorporated. The ADVANCE project facilitated continuation of this work. The motivation has been that physical network infrastructures e.g. pipelines, electricity grids, roads, railways, etc. share common characteristics in that they require high up-front investments, represent long-lived capital (stock), often display increasing returns to scale and network effects and indivisibilities and can induce considerable effects on the rest of the economy (e.g. location choices and mobility patterns induced by transport infrastructure). These characteristics imply that infrastructure networks potentially have strong implications for energy-security or climate change mitigation and therefore need better representation in IAMs. Both ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"a8leni9qa","properties":{"formattedCitation":"(Laird et al., 2005; Wilbanks, 2014)","plainCitation":"(Laird et al., 2005; Wilbanks, 2014)"},"citationItems":[{"id":37,"uris":[""],"uri":[""],"itemData":{"id":37,"type":"article-journal","title":"Network effects and total economic impact in transport appraisal","container-title":"Transport Policy","page":"537-544","volume":"12","issue":"6","source":"CrossRef","DOI":"10.1016/j.tranpol.2005.07.003","ISSN":"0967070X","author":[{"family":"Laird","given":"James J."},{"family":"Nellthorp","given":"John"},{"family":"Mackie","given":"Peter J."}],"issued":{"date-parts":[["2005",11]]}},"label":"page"},{"id":141,"uris":[""],"uri":[""],"itemData":{"id":141,"type":"book","title":"Climate change and infrastructure, urban systems, and vulnerabilities.","publisher":"Island Press","publisher-place":"[S.l.]","source":"Open WorldCat","event-place":"[S.l.]","ISBN":"1-61091-554-2","language":"English","author":[{"family":"Wilbanks","given":"Dr Thomas J"}],"issued":{"date-parts":[["2014"]]}},"label":"page"}],"schema":""} (Laird et al., 2005; Wilbanks, 2014) call for such improved representation in IAMs. From a modelling sensitivity perspective it is of interest to know what would be missed by not including physical network infrastructure in an analysis e.g. what proportion of emissions from the transport system are accounted for by its enabling infrastructure. From a macroeconomic perspective there is a large resource allocation (capital and labor) involved in the construction of physical network infrastructure which has implications for other sectors and for possible crowding out effects. The constructed infrastructure also embodies energy and has its own carbon footprint which are not insignificant as ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"2pnvsl89mn","properties":{"formattedCitation":"{\\rtf (Edenhofer et al., 2014; M\\uc0\\u252{}ller et al., 2013)}","plainCitation":"(Edenhofer et al., 2014; Müller et al., 2013)"},"citationItems":[{"id":819,"uris":[""],"uri":[""],"itemData":{"id":819,"type":"report","title":"IPCC, 2014: Summary for Policymakers. In: Climate Change 2014: Mitigation of Climate Change. Contribution of Work - ing Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Chang","publisher-place":"s, Cambridge, United Kingdom and New York, NY, USA","event-place":"s, Cambridge, United Kingdom and New York, NY, USA","author":[{"family":"Edenhofer","given":"O"},{"family":"Pichs-Madruga","given":"R"},{"family":"Sokona","given":"Y"},{"family":"Farahani","given":"E"},{"family":"Kadner","given":"S"},{"family":"Seyboth","given":"K"},{"family":"Adler","given":"A"},{"family":"Baum","given":"I"},{"family":"Brunner","given":"S"},{"family":"Eickemeier","given":"P"},{"family":"Kriemann","given":"B"},{"family":"Savolainen","given":"S"},{"family":"Schl?mer","given":"S"},{"family":"Von Stechow","given":"C"},{"family":"Zwickel","given":"T"},{"family":"Minx","given":"J.C."}],"issued":{"date-parts":[["2014"]]}},"label":"page"},{"id":166,"uris":[""],"uri":[""],"itemData":{"id":166,"type":"article-journal","title":"Carbon Emissions of Infrastructure Development","container-title":"Environmental Science & Technology","page":"11739-11746","volume":"47","issue":"20","DOI":"10.1021/es402618m","author":[{"family":"Müller","given":"Daniel B."},{"family":"Liu","given":"Gang"},{"family":"L?vik","given":"Amund N."},{"family":"Modaresi","given":"Roja"},{"family":"Pauliuk","given":"Stefan"},{"family":"Steinhoff","given":"Franciska S."},{"family":"Bratteb?","given":"Helge"}],"issued":{"date-parts":[["2013"]]}},"label":"page"}],"schema":""} (Edenhofer et al., 2014; Müller et al., 2013) have highlighted. Perhaps the biggest uncertainty with the construction and deployment of physical network infrastructure is that it can lead to lower prices and thus higher demand for end-uses. This has implications for the spatial organization of society and its greenhouse gas emissions. The approach outlined below assumes that the model being used already incorporates a transport sector and models service demand measured in passenger kilometers (PKM’s). Service demand in this sense is the distance travelled by a passenger in a particular mode. In IMACLIM-R, which is a CGE model, the general mechanism at play in this regard is that increasing personal income, leads to increased demand for transport services (measured in PKM’s) which leads to increasing energy demand. In partial equilibrium models e.g. MARKAL/TIMES, transport service demand may be exogenous. At this point in time IMACLIM-R includes infrastructure for automobiles, public transport and air travel. Freight transport or recharging and refueling infrastructure is not considered. For the infrastructure category of roads no distinction is made between motorways or other classes or roads. Public transport infrastructure is a modelling aggregation of the roads for busses and the rail tracks needed for rail travel. Further work could be to develop the categories that are not considered. The method presented below is for 12 global regions for the timeframe 2001 to 2100. The key features are that energy service demand is translated into an underlying physical infrastructure, the deployment of which is costed and constrained by physical and budgetary limits. In the accompanying excel sheet the costs of infrastructure are listed plus values for other variables which represent physical constraints, e.g. road density limits.Modelling Transport Infrastructure in the IMACLIM-R Global E3 modelTransport infrastructure in the IMACLIM-R Global model is represented by a variable called Captransport and as the name suggests it represents the available transport capacity for various modes. Captransport combines three vectors of transport modes (air, public, road) per IMACLIM-R global region into a matrix. In the model approach it is assumed that public transport (bus plus rail) has a separate infrastructure to road infrastructure. The model variable for public transport is OT (other transport). Infrastructure for freight transport is not considered. The units of Captransport are passenger kilometers (pkm). For air, although the primary infrastructure needed are airports and runways, for modelling purposes this is also grouped as a generic air transport capacity that is also measured in pkm’s. In the model to date (pre April 2015) the Captransport values of pkm were initialized for the year 2001 (calibration year of the IMACLIM-R model) by assuming that the capacity of each mode is twice the respective measured pkm’s. For example, for the Europe plus Turkey region 200 Billion pkm’s of air transport were calibrated as being flown in 2001. Thus the capacity of air transport for this region in 2001 is assumed to be 400 Billion pkm. For subsequent model years (2002 to 2100) the Captransport variable evolved at the same rate as the changing demand for transport services in the case of public and air transport and the changing stock of vehicles in the case of automobile infrastructure. This can be understood as conforming to a congestion avoidance scenario. The above is coded as follows in Scilab, a programming language similar to Matlab. The year 2001 (reference) values of PKM of infrastructure are calculated as described above for each region and placed in the Captransport Matrix:Captransport=[Capairref,CapOTref,Capautomobileref];For each subsequent scenario year the three vectors of the Captransport matrix are updated as follows: Captransport(:,1)=Captransport(:,1).*PKM_prev(:,indice_air)./PKM_prev_prev(:,indice_air);Captransport(:,2)=Captransport(:,2).*PKM_prev(:,indice_OT) ./PKM_prev_prev(:,indice_OT);Captransport(:,3)=Captransport(:,3).*stockautomobile./stockautomobile_prev;For details on how PKM and the number of cars (nb_car) in stock (stockautomobile) evolves see Waisman et al., (2013).In the model the construction and maintenance of infrastructure has happened cost free to date. In other words although varying levels of transport infrastructure were deployed each year, there were no costs assigned to this deployment.Starting in April 2015 an exercise has been conducted ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"XQlYMjoZ","properties":{"formattedCitation":"{\\rtf (\\uc0\\u211{} Broin and Guivarch, 2015)}","plainCitation":"(? Broin and Guivarch, 2015)"},"citationItems":[{"id":813,"uris":[""],"uri":[""],"itemData":{"id":813,"type":"webpage","title":"Advance Deliverable 5.3","URL":"","author":[{"family":"? Broin","given":"Eoin"},{"family":"Guivarch","given":"Céline"}],"issued":{"date-parts":[["2015"]]},"accessed":{"date-parts":[["2016",3,11]]}}}],"schema":""} (? Broin and Guivarch, 2015) as part of the ADVANCE project in which a new approach to updating the Captransport variable has been implemented. This new approach has involved, (i) adding the costs associated with transport infrastructure deployment to the model and (ii) changing the way in which the Captransport(automobile) variable evolves i.e. there is no change in the way Captransport is updated for public or air transport.. These two changes are based on an approach presented by ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"13u6sv18i7","properties":{"formattedCitation":"(Dulac, 2013)","plainCitation":"(Dulac, 2013)"},"citationItems":[{"id":212,"uris":[""],"uri":[""],"itemData":{"id":212,"type":"article-journal","title":"Global Land Transport Infrastructure Requirements: Estimating road and railway infrastructure capacity and costs to 2050","container-title":"International Energy Agency, Paris.","journalAbbreviation":"International Energy Agency, Paris.","author":[{"family":"Dulac","given":"John"}],"issued":{"date-parts":[["2013"]]}}}],"schema":""} Dulac, (2013) for the IEA. Dulac’s costs are used in this work to provide calibration values for the cost of infrastructure for the model calibration year, 2001. Costs for road infrastructure are made up of new roads and parking spaces, upgrade of existing roads and O&M for existing roads and parking spaces. Costs for construction and O&M of Air infrastructure (airports) have been estimated independently using data from the ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"tlEtlabx","properties":{"formattedCitation":"(OECD, 2015)","plainCitation":"(OECD, 2015)"},"citationItems":[{"id":801,"uris":[""],"uri":[""],"itemData":{"id":801,"type":"webpage","title":"Statistiques OCDE","URL":"","author":[{"family":"OECD","given":""}],"issued":{"date-parts":[["2015"]]},"accessed":{"date-parts":[["2015",12,23]]}}}],"schema":""} (OECD, 2015). The accompanying spreadsheet, advanceW5.4_data.xlsx, lists the dataset of transport infrastructure costs in $2010 prices as applied in this work. As IMACLIM-R is calibrated to year 2001 a price deflator ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"2eahclfmpa","properties":{"formattedCitation":"(Friedman, 2015)","plainCitation":"(Friedman, 2015)"},"citationItems":[{"id":881,"uris":[""],"uri":[""],"itemData":{"id":881,"type":"webpage","title":"The Inflation Calculator","URL":"","author":[{"family":"Friedman","given":"Morgan"}],"issued":{"date-parts":[["2015"]]},"accessed":{"date-parts":[["2016",9,14]]}}}],"schema":""} (Friedman, 2015) is used to adjust the costs given in advanceW5.4_data.xlsx. accordingly. In IMACLIM-R costs are made up of price and volume. For the model calibration year (2001) the price of construction (for all sectors) is set to 1 while the volume is a dimensionless unit of construction, equal in absolute value to total investment in construction (infrastructure plus other construction investments). The price, as with the prices of in other sectors, evolves in the IMACLIM-R in response to the macroeconomic components of the model. Because both the prices and the size of the infrastructure evolve, the costs of infrastructure evolve too. Thus while the costs of infrastructure are defined for 2001 (see previous paragraph) they are scaled for each subsequent year by an index which represents the change in prices of the construction sector. For the costs some adjustments were needed. For example the costs for public transport seemed initially to be too high. This was found to be because they are for rail whereas the IMACLIM-R public transport variable also includes bus transport which is relatively cheap. Therefore it was decided to multiply the costs for public transport construction by the share of rail in public transport (see parameter 7 in Table 2 – data from ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"1vp71p5g4u","properties":{"formattedCitation":"(Dulac, 2013)","plainCitation":"(Dulac, 2013)"},"citationItems":[{"id":212,"uris":[""],"uri":[""],"itemData":{"id":212,"type":"article-journal","title":"Global Land Transport Infrastructure Requirements: Estimating road and railway infrastructure capacity and costs to 2050","container-title":"International Energy Agency, Paris.","journalAbbreviation":"International Energy Agency, Paris.","author":[{"family":"Dulac","given":"John"}],"issued":{"date-parts":[["2013"]]}}}],"schema":""} (Dulac, 2013)). To allow for ongoing modernization of public transport systems, e.g. the deployment of trams, it was assumed that the share of rail in public transport increases by 0.5% per annum in seven of the twelve IMACLIM-R global regions (China, India, Brazil, Middle East, Africa, Rest of Asia and Rest of Latin America). The main purpose of this assumption is to reflect the increased costs of public transport deployment over the scenario period. The following lines of code show how costs for roads are applied. The costs themselves are listed in columns B of sheet Costs in the accompanying excel file:C_new_roads=cost_road_constr.*New_roads.*conv_pkm_lanekm;C_oandm_roads=cost_road_oandm.*0.25.*Captransport(:,3).*conv_pkm_lanekm;C_upgrade_roads=cost_road_upgrade.*0.05.*Captransport(:,3).* conv_pkm_lanekm;C_parkings_construct=park_space.*(max(0,(nb_car-nb_car_prev)).*cost_parking_constr);C_parkings_upgrade=nb_car_prev.*0.05.*cost_parking_upgrade;C_parkings_onm=nb_car_prev.*0.33.*cost_parking_oandm;C_parkings=(C_parkings_construct+C_parkings_upgrade+C_parkings_onm);C_roads=C_new_roads+C_oandm_roads+C_upgrade_roads+C_parkings;The first line shows the costs of new road construction (per lane-km) being multiplied by the length of new roads (in pkm) using a conversion factor for the different units (conv_pkm_lanekm). The scalars e.g. 0.25 reflect how often a cost is applied, i.e. every four years in the case of cost_road_upgrade. Note that if the stock of cars decreases no new parking spaces are constructed but at the same time parking spaces are not removed either. The costs per region of construction, upgrade and operation and maintenance i.e. cost_road_constr, cost_road_oandm, cost_road_upgrade, cost_parking_constr, cost_parking_upgrade, cost_parking_oandm, and the size of parking spaces (park_space) are listed in AdvanceW5.4_data.xlsx. It is assumed that for every car in stock there are either two or there parking spaces and that their sizes are either 15m2 or 18m2. Two spacese are necessary as each vehicle must have it’s current location and a possible destination. The differences given are region specific.The following section describes how the length of new road and is calculated and thus how Captransport(:,3) is updated.For roads (i.e. used for automobiles but not public transport) the change in Captransport(:,3), referred to below as Captransport(automobile) , in remodeled to occur due to the following five constraints, of which numbers 3 to 5 are new model developments:The utilization rate of the road network.The change in the stock of vehicles.The construction capacity in the region. See parameter 2 in REF _Ref437880080 \h \* MERGEFORMAT Table 1.The density of the existing road network. See parameter 3 in REF _Ref437880080 \h \* MERGEFORMAT Table 1.The maximum percentage of GDP that can go to infrastructure The first point emphasizes that existing roads may be underutilized and thus an increasing number of vehicles on the road or KM’s driven does not necessarily mean that new roads are needed. The second point reflects that an increasing stock of vehicles can lead to increased travel use and pressure to build infrastructure. The third point seeks to incorporate the limits of the construction industry itself i.e. that a ramp-up in levels of construction can only happen if the requisite labor and capital resources, and technical knowhow exist. The fourth point provides a realistic alternative to scenarios of linear growth of infrastructure capacity. In such linear growth scenarios the density of road infrastructure in India can reach the same level as that of Manhattan, New York, by 2050, a clearly implausible outcome ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"1oej8cbq3d","properties":{"formattedCitation":"(Dulac, 2013)","plainCitation":"(Dulac, 2013)"},"citationItems":[{"id":212,"uris":[""],"uri":[""],"itemData":{"id":212,"type":"article-journal","title":"Global Land Transport Infrastructure Requirements: Estimating road and railway infrastructure capacity and costs to 2050","container-title":"International Energy Agency, Paris.","journalAbbreviation":"International Energy Agency, Paris.","author":[{"family":"Dulac","given":"John"}],"issued":{"date-parts":[["2013"]]}}}],"schema":""} (Dulac, 2013). The final point emphasizes the average percentage of GDP that has been spent on infrastructure to date. The data for these constraints is listed in the Captransport Constraints sheet of the accompanying excel file.The constraints are implemented as follows. It is assumed that each region is striving for a roadway utilization rate of 50% (UR_automobile_ideal). Utilization rate in this sense is a modelling construct whereby the number of pkm’s travelled is divided by the infrastructure capacity (also measured in pkm’s) and measures levels of congestion. 50% is chosen as an average between the current high levels of utilization (90%) in Brazil i.e. a high level of road congestion and low levels (15%) in India i.e. a low level of road congestion. The evolution of pkm is anticipated to follow the previous year increase (or decrease) in stock of vehicles and average pkm driven per vehicle. In parallel it is anticipated that the number of pkm’s driven for each new year (pkmautomobile_anticip) changes relative to the annual change in the stock of vehicles (itself related to changes in the level of income) and the average pkm’s driven per vehicle in the previous year. A combination of this anticipated pkm increase and the target utilization rate (as described above) gives the planned Captransport(automobile) for the subsequent year as follows:capautomobile_target=pkmautomobile_anticip/UR_automobile_target This result is then compared against the third and fourth constraint listed above : (i) The construction industry capacity in the region (Parameter 2 in REF _Ref437880080 \h \* MERGEFORMAT Table 1), and (ii) The density of the existing road network (Parameter 3 in REF _Ref437880080 \h \* MERGEFORMAT Table 1). These are two checks on the amount of new road infrastructure that the model estimates. The construction capacity is assumed to change annually with the increase or decrease in production in the construction sector. Annual production for each sector is calculated in the static equilibrium (see above) of the model. The density limit is assumed to be constant. A variable New_roads is then defined as the difference between Captransport(automobile) from the previous year and that calculated according to the aforementioned constraints. New_roads is then added to Captransport(automobile) from the previous year to update this metric. A final constraint on Captransport(automobile), the fifth listed above, the amount of investment that can go on infrastructure (Max_Infra_Road_Invest), is then introduced. This has initially been set at 2% of the value of GDP in accordance with a recently published ITF report ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"1ntfp9tc9j","properties":{"formattedCitation":"(OECD/ITF, 2015)","plainCitation":"(OECD/ITF, 2015)"},"citationItems":[{"id":823,"uris":[""],"uri":[""],"itemData":{"id":823,"type":"report","title":"ITF Transport Outlook 2015","publisher":"OECD Publishing/ITF, ","URL":"","author":[{"family":"OECD/ITF","given":""}],"issued":{"date-parts":[["2015"]]}}}],"schema":""} (OECD/ITF, 2015). The report states that road spending has been generally found to decline with the level of GDP per capita. For OECD countries the level is currently around, 1%, while for eastern European and Asian countries it is 2% or higher. The 2% cap covers the cost associated with road infrastructure (construction, upgrade, O&M and parking spaces). Given the substantial road networks in place in OECD countries upgrade, O&M and parking spaces (Parameter 6 in REF _Ref437880080 \h \* MERGEFORMAT Table 1) can combined make up over 50% of spending on road infrastructure. Model testing has shown that this constraint also prevents the capacity of road transport rising too fast. Despite the aforementioned constraints this can still occur in the model. This is because the increased capacity modelled for Captransport allows greater distances to be travelled (pkmautomobile) within the travel time budget (as modelled in the static equilibrium of IMACLIM-R), and thus pkmautomobile_anticip (see above) which is the basis for capautomobile_target, can increase rapidly. If it is found that the combined cost of roads, C_roads, exceeds 2% of GDP, the variable New_roads is recalibrated to be the length of road that would be possible to construct for the difference between 2% of GDP and the combined cost of road O&M, upgrade and parking spaces. Table SEQ Table \* ARABIC 1 : Modelling parameters for road infrastructure in IMACLIM-R Global model.Parameter NameValueUnit1UR_automobile_ideal50% pkm/Captransport2constr_limit30 – 355a Lane-km/year in thousands3density_limit1-6aLane-km per km2 land4conv_pkm_lanekm4-5 X 10-7Lane-km/pkm5Max_Infra_Road_Invest2% of GDP6park_space2X15 – 3X18aSquare Metres where 2x means two parking spaces 7share_rail_OT5 - 70 a%aValues for model reference year which vary depending on regional circumstances Note that the budget cap does not cover public transport or airports as infrastructure costs for roads make up 70% or more of inland infrastructure investment ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"1rls0ndc34","properties":{"formattedCitation":"(OECD/ITF, 2015)","plainCitation":"(OECD/ITF, 2015)"},"citationItems":[{"id":823,"uris":[""],"uri":[""],"itemData":{"id":823,"type":"report","title":"ITF Transport Outlook 2015","publisher":"OECD Publishing/ITF, ","URL":"","author":[{"family":"OECD/ITF","given":""}],"issued":{"date-parts":[["2015"]]}}}],"schema":""} (OECD/ITF, 2015), but also because doing so would necessitate a decision as to how to allocate budget between road, public transport and air. In an alternative scenario other budget caps on spending for public transport or airports infrastructure could be introduced.Thus to summarize, Captransport (automobile), increases to ensure that congestion is avoided despite increasing numbers of vehicles and pkm’s driven, but within the bounds of a realistic level of construction set by, the construction industries capabilities, the existing road density and the budget available for road infrastructure. REFERENCES ADDIN ZOTERO_BIBL {"custom":[]} CSL_BIBLIOGRAPHY Dulac, J., 2013. Global Land Transport Infrastructure Requirements: Estimating road and railway infrastructure capacity and costs to 2050. Int. Energy Agency Paris.Edenhofer, O., Pichs-Madruga, R., Sokona, Y., Farahani, E., Kadner, S., Seyboth, K., Adler, A., Baum, I., Brunner, S., Eickemeier, P., Kriemann, B., Savolainen, S., Schl?mer, S., Von Stechow, C., Zwickel, T., Minx, J.C., 2014. IPCC, 2014: Summary for Policymakers. In: Climate Change 2014: Mitigation of Climate Change. Contribution of Work - ing Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Chang. s, Cambridge, United Kingdom and New York, NY, USA.Friedman, M., 2015. The Inflation Calculator [WWW Document]. URL (accessed 9.14.16).Laird, J.J., Nellthorp, J., Mackie, P.J., 2005. Network effects and total economic impact in transport appraisal. Transp. Policy 12, 537–544. doi:10.1016/j.tranpol.2005.07.003Müller, D.B., Liu, G., L?vik, A.N., Modaresi, R., Pauliuk, S., Steinhoff, F.S., Bratteb?, H., 2013. Carbon Emissions of Infrastructure Development. Environ. Sci. Technol. 47, 11739–11746. doi:10.1021/es402618m? Broin, E., Guivarch, C., 2015. Advance Deliverable 5.3 [WWW Document]. URL (accessed 3.11.16).OECD, 2015. Statistiques OCDE [WWW Document]. URL (accessed 12.23.15).OECD/ITF, 2015. ITF Transport Outlook 2015. OECD Publishing/ITF, , D.T.J., 2014. Climate change and infrastructure, urban systems, and vulnerabilities. Island Press, [S.l.]. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download