You are on page 1of 10
Urban Analytics and City Science 2017 Volume 44 Number 3 May ISSN 2399-8083 Environment and Planning B | STAFFS UNIVERSITY LIBRARY b i journals.sagepub.com/hom: | 552 Environment and Planning B: Urban Analytics and City Science 44(3) if not arepy.Exists(in_Program): raise Exception, “Input program does not exist” # Run the model / program arcpy.AddMessage(“*) arepy.AddMessage(“Running %s” % (in_Program)) dese = arepy.Deseribe(in_Program) sourceFilePath = dese-path os.chdir(sourceFilePath) os.system(in_Program) # Export shared files sourceFilePath = in_Scenario skm2SILO1=sourceFilePath+"\\HwyPK_iter6.skm” skm2SILO2=sourceFilePath-+"\\WTmPK.skm” # Convert Skim Matrices from the MSTM to OMX Matrices for SILO cube2omx_converter = “C:\\models\\cube2omx\\cube2omx.exe” os.system(cube2omx_converter+ “ “ + skm2SILO1) os.system(cube2omx_converter+ “ “ + skm2SILO2) omx2SILO1=sourceFilePath+”|\HwyPK_iter6.omx” omx2SILO2=sourceFilePath+"\\WTrnPK.omx” # Rename output files to include the modelling year as defined in SILO properties file omx2STLOI_year=sourceFilePath-+”\\HwyPK_iter6_""+str(in_Year)+”.omx” omx2SILO2_year=sourceFilePath+"\\WTmPK_"+str(in_Year)+".omx” if arepy.Exists(omx2SILO1_year): os.remove(omx2SILO1_year) os.rename(omx2SILO1, omx2SILO1_year) if arepy.Exists(omx2STLO2_year): os.remove(omx2SILO2_year) os.rename(omx2SILO2, omx2STLO2_year) arepy.AddMessage("“Exporting exchange data files:”) arcpy.AddMessage(omx2SILO1_year) arcpy.AddMessage(omx2SILO2_year) arcpy.SetParameter(3, omx2SILO1_year) arcpy.SetParameter(4, omx2SILO2_year) elapsed = (time.time() - start) arcpy.AddMessage(“) arepy AddMessage(“Model end time: Ys” % timestrftime(‘%X %x %Z")) # Handle script errors except Exception, errMsg: ; # If we have messages of severity error (2), we assume a GP tool raised it. # Otherwise, we assume we raised the error and the information is in errMsg. if arepy.GetMessages(2): arcpy.AddError(arepy.GetMessages(2)) else: arcpy.AddError(str(errMsg)) a Article Environment and Planing & Urban ‘Analytics and City Science 2017, Vol 44() 553-569 © The Author(s) 2016 Reprints and permissions sagepub.coulijourrlsPermisions.nav Urban infrastructure is not a tree: Integrating and decentralizing urban ot 10 77aaselStecerae? infrastructure systems aon SAGE Sybil Derrible ‘Complex and Sustainable Urban Networks (CSUN) Laboratory, University of linofs at Chicago, (Chicago, USA Abstract In his original 1965 article, Christopher Alexander argued that master planned cities ultimately failed because the designs elaborated followed a tree structure as opposed to a more desirable semilattice structure present in organic cities. In this article, | argue that a similar claim can be ‘made with urban infrastructure systems planning. As cities expanded and became increasingly complex in the 20th century, the responsibility to plan and design urban infrastructure was distributed to separate agencies that seldom communicate and coordinate with one another. In the global context to make cities more sustainable and resilient, a better integration of infrastructure systems may hold much potential. After recalling Alexander's main concepts, | examine how current infrastructure systems are naturally interdependent. | then discuss the role of integration, by notably proposing an integration-decentralization matrix, with four quadrants, illustrated by using practical examples. The quadrants are current paradigm, siloed distribution, localized integration, and integrated decentralization. Overall, a better integration of urban infrastructure can oer significant benefits to a city, and it may be time to seriously revi ‘our current urban infrastructure systems planning practice. Keywords Urban infrastructure planning, sustainability, resilience, integration, decentralization Introduction In his seminal 1965 article, 4 City is not a Tree, Christopher Alexander argued that the elements within cities that grew organically are highly interconnected, in contrast to master planned cities (A la Le Corbusier), in which elements are physically separated, following what Alexander refers to as a tree network structure (Alexander, 1965). Alexander was referring to elements of urban form as opposed to urban infrastructure systems (UIS) that include water Corresponding author: ‘Sybil Derrible, University of tlinois at Chicago, 842 W. Taylor Sereet (M/C 246), Chicago, linois, 60607, USA. Email: derrible@ulcedu 554 Environment and Planning B: Urban Analytics and City Science 44(3) conduits and electric lines that are most often not visible (with the exception of transportation infrastructure), but his lessons may be relevant for UIS planning and design as well. After all, streets do not only carry traffic, they also carry water runoff, and urban flooding is becoming an increasingly significant problem in cities (Dawson, 2007; Kermanshah et al., 2015; Kundzewiez et al, 2013). Moreover, with the global environmental crisis, the pertinence to study and better plan UIS is particularly timely since the energy and resources used in cities are systematically distributed through UIS, from water, electricity, and gas, to rail tracks and roads (Liu et al., 2015; Mohareb et al., 2015). In addition, there has been a recent push to better understand how UIS are integrated for resilience purposes, further highlighting the need to better plan and design UIS (DHS, 2015; Duefias-Osorio et al., 2007; Duefias-Osorio and Vemuru, 2009). Indeed, because of their inherent integration, the failure of one UIS can cascade to another UIS, and a more coordinated and better planned integration is highly desirable, in contrast to the current state of the practice that Sage et al. (2015) describe as being “oriented around a short-term, stable system (STSS) perspective.” ‘The need to better physically integrate UIS is not new. In fact, UIS used to be integrated up to the late 19th century and early 20th century as city engineers and chief architects were often responsible for the planning and operation of all UIS; Mumford (1961; 172) describes them as “regimenters of human functions and urban space.” For instance, Baron Haussmann designed both Paris’ sewer system as well as Paris’ main arterials to accommodate traffic. ‘Neuman and Smith (2010) offer an interesting account of the historical relationship between city planning and UIS planning, notably recalling that the term “infrastructure” itself may only date back to 1875 France. As cities expanded, however, the responsibilities to run each UIS were divided among various departments; for instance, at the national level, the UK Ministry of Transport was created in 1919 (National Archives, 2015), and the US Department of Transportation was only created in 1966 (US DOT, 2015). Similarly, the US Department of Housing and Urban Development was only created in 1965 (HUD, 2015). In particular, the global push toward safety, accountability, and higher efficiency further encouraged this dichotomy of responsibility. This is not to say that having separate departments is undesirable, UIS are complex, and they cannot be handled by one common entity (as well argued by Minett (1975)—partially in response to Alexander's article), but this separation of responsibility also inhibited communication and coordination between departments. ‘Consequently, many recognize the need for an alternative form of urban infrastructure planning practice, pushing for the design of UIS as integrated systems; this is sometimes referred to as “infrastructure ecology” (Pandit et al., 2011; Xu et al., 2012), Indeed, adopting aan integrated design of UIS is especially relevant today with the global movement toward decentralizing and distributing major infrastructure systems (Lee et al., 2013; Makropoulos and Butler, 2010; Pagani and Aiello, 2011; Pahwa et al., 2013), and supported by the relatively recent push toward higher resilience (Fiksel, 2003; Holling, 1973; Perrings, 1998; Woods, 2015), but the focus of this work is not on resilience. Here, integration is understood in a broad sense, including both the sharing of physical space (e.g. water conduits in streets) and the sharing of resources from an urban metabolism viewpoint (c.g. electricity needed to distribute water). There therefore seems to be a natural parallel between Alexander’s ideas on planning and urban form with the current shift toward UIS integration. The main goal of this article will be to study this parallel. More specifically, the objectives of this work are to: ‘© Recall the contribution and concepts of A City is not a Tree; Analyze the current interdependencies between UIS; © Discuss the role of UIS integration in the future, supported by a conceptual matrix and notable case studies. eI Derrible 555 al Tree Network Semi-Lattice Figure I. Network representation of (a) tree versus (b) semilattice. In this article, each objective is tackled sequentially. In the next section, I will therefore recall the main concepts that Alexander presented in A City is not a Tree. This section is followed by a discussion on how current UIS are integrated, which will include a table showing some of the ways UIS are interdependent. Finally, I will expose some thoughts on how these current interdependencies can be better planned and coordinated in the future, by notably considering the importance of the current movement for decentralizing and distributing UIS, ‘which will be supported by a conceptual matrix and illustrated by several real-world examples. Overall, the contribution from this work possess much potential, especially if integrated with complexity science that shows great promise to study urban systems (Derrible and Kennedy, 2011; Ahmad and Derrible, 2015; Derrible and Ahmad, 2015; Peiravian et al., 2014). Alexander's A City is not a Tree Despite having been written more than half a century ago, the contribution of Alexander remains pertinent today. In fact, it can be expanded to many disciplines beyond architecture and urban planning. His main argument is that elements in organic cities do not follow a tree structure, By tree structure, what is meant here is the mathematical structure of a tree, which translates into physical delineation or a division that prevents any elements from “belonging” or “overlapping” in two or more realms (i.e. physical spaces). Figure | shows the example of a tree structure on the left relative to a semilattice structure on the right that allows overlapping and crosscutting. To be exact, a tree is a special kind of semilattice; one that does not allow any overlaps. In a tree structure, all d-elements belong to one single c-element, which in turn belong to one single b-element and so on. In a semilattice, however, and despite keeping the same overall hierarchy, all elements can be connected, Toillustrate the concept, Alexander discusses the example of a newsstand in front of a store that will attract the eye of pedestrians when waiting to cross a street at a traffic light. The newsstand is therefore connected to both the store and the traffic light (as opposed to being only connected to the store). Another way to understand the overarching concept is to think of how neighborhoods are often divided in cities. Many cities tend to have older neighborhoods located next to one another and that are considered different from an administrative perspective, but the urban form between them is seamless, thus connecting them. In contrast, suburban areas often have highways or large arterials that clearly delineate their boundaries. A seamless integration allows for elements to be part of multiple neighborhoods, while a strict separation does not. Interestingly, at the time of publication of this article, DNAInfo (2016) was organizing an online program for New York residents to draw the boundary of their neighborhood (responses were not published, however) 556 Environment and Planning B: Urban Analytics and City Science 44(3) Derrible 557 ; through UIS; buildings also often “compete” with UIS for space (ie. land availability). In Tree Network. Semni-Lattice this work, UIS therefore consist of: Figure 2. Venn representation of (a) tree versus (b) semilatice. Another way to look at the examples given in Figure 1 is to consider Venn diagrams as opposed to networks. In Figure 2, the diagram on the left reproduces the tree network from Figure 1, while the diagram on the right reproduces the semilattice. Thanks to its symmetry, the tree structure is much easier to understand and interpret from a cognitive viewpoint. It also “looks” more orderly than the semilattice. At the design phase, the tree network may therefore seem preferable than the semilattice; common epithets used in urban design include “elegant” and “streamlined,” which would support this argument for a seemingly more orderly design with a tree network structure, In fact, what Alexander argues is that humans have a propensity to design neighborhoods that have a tree structure, whereas successful neighborhoods that have evolved over time follow a semilattice structure. In fact, Alexander goes further and argues that human beings simply cannot design semilattices because of the way the brain functions. This delineation or division process would therefore result from a cognitive inability to produce semilattices directly. The purpose of this article is not to validate or refute this idea, but to understand how it applies to UIS planning. With notable exceptions, I would argue that a similar tree structure has been applied in how UIS are planned and operated, dichotomizing them as individual entities as opposed to being part of a larger urban system. Moreover, the division of administrative responsibilities in separate (and even competing) departments has produced a tree-like structure. Some departments are responsible for the water system, while others are responsible for the transportation system. We now commonly hear of examples where a road is dug and repaved one year to change a water conduit, and it is also dug and repaved the following year to change a gas line. Keeping the example of streets, in our current paradigm, streets are designed by transportation engineers who apply rules from their geometric design handbooks, with an emphasis on accommodating traffic. As mentioned previously, however, streets must now be designed differently and have more green infrastructure to accommodate surface runofis (e.g. by using rain gardens—discussed later), but the process to systematically include green infrastructure has been painstakingly slow. This problem spans beyond the example of streets, however, and in the next section, I will discuss how UIS are inherently interdependent. Interdependency between UIS UIS can include many different types of infrastructure. For this work, I solely include the UIS that are directly planned by a municipality or that use municipal space. Moreover, I include buildings as the end points where the energy and resources are being distributed © transportation: roads, rail tracks, sidewalks, and back alleys; © water: conduits for distribution, pumps, stormwater channels, stormwater and sanitary sewers lines; © utility: mainly gas but utility can include steam or chilled water for district heating and cooling systems; © electricity: primarily distribution lines in cities but also small transmission lines (roughly up to 50k); © telecom: phone lines and internet cables; © buildings: in the form of land use and demand points for the resources carried by the other UIS, Although they are typically considered a utility, water and electricity will be considered on their own in this work since their properties are vastly different and they cannot be grouped with other utilities. Table 1 contains a nonexhaustive list of ways that each UIS is integrated with one another. The table reads as follows. The rows describe how each UIS is linked or impacted by the UIS in the columns. The first row therefore describes how transport is affected by water, utility, electricity, telecom, and building infrastructure. For example, streets are being used to locate water conduits, gas, and telecom lines. Transport requires electricity for electric vehicles, electric rail and bus modes, and to operate traffic signals and streets lights. Finally, buildings often dictate how transport infrastructure is used as origins and destinations of most trips. While I will not repeat in the text the information given in the table, I will discuss the level of integration of each UIS separately. Transport Physically, transport infrastructure is highly integrated with most UIS since water conduits and sewers, electric wires, gas, and telecom lines tend to be located in streets, either elevated, at grade, or underground. This can either be seen as competition for space, but it also offers opportunities for better integration. The fact that these UIS are located in streets also suggests that they share a common spatial distribution, and they may therefore share many spatial properties. This can be significant in the current movement toward distribution and decentralization. Moreover, transport infrastructure most often carries the raw materials needed for the other UIS, from carrying fuel for power generation and products needed for water treatment, to handling the disposal of outputs from UIS such as the biosolids from wastewater treatment. The integration of transport with buildings is also natural since both systems are complementary. Recently, the integration between the water and transport systems has become more preeminent. As mentioned above, urban flooding is becoming an increasing problem During heavy rain events, and because most surfaces in urban settings are impervious, rain is directed to catchment basins at street curbs, and stormwater infrastructure rapidly becomes incapable to handle these large flows of water. This is especially a problem for combined sewer systems that need to handle both sanitary and stormwater sewers, frequently leading to combined sewer overflow (dumping of untreated water in natural streams). Although the overflow of separated sewer systems generates fewer negative impacts (because stormwater is not as contaminated), flooding remains a significant problem. 559 Derrible Environment and Planning B: Urban Analytics and City Science 44(3) 558 uonisinbsy seq pur janue> Aiosiasedns ‘ya¥Os dumonseayuy eumanaseajut $e Je8se| 40) aren 198124 405 Pur] 10} 221y40D © pur] 40} 2>JUOD souy uo|sssueD, Buyjooo (ouausaseg pinsip Bujpooy 02 pea} . . swasis © — suiojgoud sory, © $280 Jaren jo (sreiso yeou ur 0) Anqeyene Aanigeyene ‘uone>9} Supjing © . . ppouasaig © PRDUDS01g © PUR 3O}229U0D © upyng spe jes 02 ax0u Suyooo 40 pe1e20} aue 501929 ‘Axpun2e[@ unowe syo1ue> veusoauy Aue ueoytuBis e eunbas exep ur Ayenadsa punou8iopun smewuas meq « “Buyoo> 40} pare0} payreasur Apunsepe auinbou Papeou ose Jovem ——_ualYA SOUN| LHO29F—2 axe seuy) wos9yo seo1cap worojaa ty © jo siunowe 08] © yeas or puty ‘au0yM sulod a2eds puno.Ziepun __aseds punoufiepun _aveds punouBsepun fom, puo se sSuppyng « Jojuopnedwod © — Jojuopnedwoy © — Jojuonneduoy © joryBU PaDINSY © — esp opeseue8 Apunseye sBupyng 4oy euerews on axou soypueag ‘er Jo 1WaUIEMO,] © pan (nim puezey © (snxou sarem punouBrepun uonduinsuo> sore ews porto] Aaounoaye 30 {ym wos9fea uo se yeamneu nse seseid se sBuping © soumjou Buiseo.5u, ‘wosy uonesoue8 pejjensur ose souy ‘roreM 02 Jes ® pune © — samod yeuuoyy © . vonnginsip moy —_apeds punouBrepun aoeds punouZropun _aoeds puno.iBsepun ‘fem jo sieup Ayenstg © 40) uonnadwos Joy uonneduioy * — Joyuonnadwioy © au painsey © bouee supa woos a “on eM weds “penunues “| 21qeL (penunuo>) punouBsopun Buyooo pue pare2o} Suneay eneds 40y Buponuow voy sedid seven uondunsuo> s2em auup-yead pejyy> pu wieens Jo sored se sBuping © vvonduinsuo> se8 jo 40} vorssiusuen uopeuoju] « ‘apeds punos2iepun ‘oeds punosBsapun aoeds punosBiepun sored se sBuping © — Jojuopnadwo © 40} uonnedwo> + Joy uonmadwioy © — YB eLNSeY © —fymaq suarour arews Bupooy vonduunsuo> ‘gy wo39jaa uo (01 Bupeay saoeins sorem jo oaueyjor Suyseeuouy © aiqeouoduyy © seoeid se sBuping sjjom Joxempunous punoufrapun Bupooy (snxou sorem reulure3u0> art29} on Burpea| saaejins AB rata) 101m ued 9e3] 38D vay aummn.nsesjut aiqvousoduy © ‘ainginsip pue ——_uonngunsip so1e nem siea.ns mojeq cr uomeWoj) © anor Apinsag © 40} sdund unssey © wpeos 03 pie * ‘qn sunpuos ——aoeds punouBiepun aaeds punouiepun ——_apeds punoalepun fem yewem aoiog © oj uonnadwo> © — Jojuonnedwo> © 40} uonnadwo> « JOrBU PEPINSY © — sung Buipooy suonesedo 403 pue uondnusip 03 Buypeoy sjouuey> ‘sapour sng pur jres yen on Supe} seremunions DUND9IE “S9IIIYOR Jo BuMOysaAQ © 219999 40} Supooy, oy asedep popeou Aapunsag © soanns 03 3uppea| 40 01 08 ajdoad . wonesoues © syouns pur seq # a1ay uone29| Aapinsaie swans se sfuping © uy souy woo9je soy uodsuen, ‘ul sunpuoa s21em pur] toy 22) # puno.Brapupy © Feuoreur Mey © . punoiBiopun fi sBupiea wo3er oo “oxen = “vodsvea, “sepuapuadapseruy Jo Asouaau! SIN “1 O1GeL 560 Environment and Planning B: Urban Analytics and City Science 44(3) Water ‘After transport, water infrastructure systems tend to be the oldest in a city. From the large water infrastructure that would bring clean water to a city during the Roman times (what Sedlak (2014) calls Water 1.0), large water conduits have been buried in streets since at least the late 19th century when steam power could provide the pressure necessary to distribute the water. In fact, some of the water conduits can at times be so old that their carrying capacity has become insufficient to service the current demand. Moreover, the presence of leaks can have substantial impacts, requiring significant inereases in energy (i.e. electricity and gas) to distribute the water. In the US, leaks account for about 8% of the total consumption (Chin, 2014), and in London, leaks were reported to account for up to 28% of all water consumed (Kennedy et al., 2007). The water system is therefore highly reliant on the provision of constant power to be able to function. In fact, pumping alone typically accounts for about 80% of all energy consumed in water distribution systems (Sedlak, 2014). The only exception is for gravity systems, in which water is stored above all consumers (c.g. on a hill or in a water tower), thus using gravity to maintain a certain pressure. Nevertheless, water treatment still requires power. Water is also sometimes integrated with gas lines, since water distribution pumps can be powered by gas; in 2016, four of the 12 pumping stations in Chicago were powered by gas but some stations are being converted to electric. In the age of smart meters, the telecom UIS is also pervading in water. Telecom has already been present for quite some time (since the 1960s) to monitor pressure at mains through common Supervisory Control and Data Acquisition (SCADA) systems, and this integration is likely to increase in the future, Moreover, buildings and water are naturally integrated since every building is directly connected to the water system Utility As defined here, the utility systems include mainly gas systems in urban settings. Gas is used in cities for two primary reasons: heating and cooking, and it can also be used for water heating. Akin to water, an extensive network of gas lines populates the underground of most cities to distribute gas to consumers. Also akin to water, SCADA systems monitor in real time how gas is being distributed in a network. Because not all buildings use gas, however, not all buildings are connected. In fact, appliances that run on gas can be substituted with electric appliances, and new neighborhoods and cities have the choice to opt out of building a gas system, which is arguably desirable since gas distribution systems can also present severe safety issues in the event of leaks. Moreover, except for the fact that they are also located underground (and gas is sometimes used for water distribution), the gas system is arguably the least integrated infrastructure with other UIS. Utility also includes district heating and cooling systems (Rezaie and Rosen, 2012). These latter systems are pipe systems that distribute steam in the winter to heat up buildings and chilled water in the summer to cool down buildings. The main premise is that instead of equipping every single building with furnaces and chillers, district-wide systems can benefit from economies-of-scale type savings. These systems are relatively common in central business districts, sometimes using local resources as the deep lake water cooling system in Toronto (Fotinos, 2003). In fact, district heating and cooling systems possess a great potential to reduce overall building energy demand. But because of relatively high thermodynamic losses, the size of these systems remains relatively small. This can be an asset, however, when thinking of decentralized systems. Moreover, the scale is particularly suited for combined heat and power (or cogeneration) systems (IEA, 2014). Derrible 561 Electricity If water is essential to life, electricity has arguably taken the most preeminent role in our societies. Nearly everything we do requires some electricity. Keeping a robust system has therefore become a priority. Accordingly, the electric grid is deeply integrated with the other UIS. Akin to water and utility, electric wires follow street patterns, either on posts or buried underground. The latter is often preferable since tree branches present a serious hazard for electrical systems; in fact, power distribution companies often allocate a significant budget to tree branch trimming. Perhaps more importantly, the electric system also often relies on transport UIS for the provision of primary energy (ic. gas, nuclear, coal) to generate electricity, although the scale here is often regional as opposed to urban. In contrast, renewable energy harvests local resources, but easy access to renewable energy facilities is required for maintenance purposes for instance, which can become a problem since the surface area covered by solar and wind farms is significantly larger than the area covered by conventional thermal electric power plants. The electric system is also highly integrated with the water system, since thermal systems need massive amounts of water for cooling. This is typically referred to as the “energy-water” nexus and it has been studied heavily (Bazilian et al., 2011; Byers et al., 2014; Kenway et al., 2011; Qin et al., 2015; Sovacool and Sovacool, 2009), although one may argue that the streams of water used for this purpose are not part of the urban water infrastructure system. The electricity and gas systems do not tend to be well integrated except for their spatial distribution. That being said, beyond the physical UIS, how electricity and gas are consumed can be related. For instance, in cities with high temperature gradients between summers and winters, electricity consumption tend to be higher in the summer for space cooling (Derrible and Reeder, 2015), while gas consumption tends to be higher in the winter for space heating (hence relieving the need for electricity); this is why major power outages tend to happen in the summer. Furthermore, and akin to water, the integration between electricity and the telecom infrastructure is increasing rapidly with the growing use of smart meters. Also akin to water, the electric grid is deeply integrated with buildings as consumers of electricity. Finally, unlike other UIS, the current electricity generation paradigm is changing, from having a one-to-many network structure to having a few-to-many structure in this age of deregulation (Larsen and Arango, 2013), and it could even lead to a many-to-many type structure with the increasing installation of solar panels and urban scale wind turbines (more on this in the next section). Telecom The telecom industry has boomed since the rise of the Internet. While telecom used to only consist of the telephone, it has since been overtaken by Internet companies that consistently lay out fiber optics cables in streets all around the world (Blum, 2012). In fact, many new cables are placed in old pipes from the Western Union in the US. Beyond the streets, the telecom infrastructure is well integrated with the transport system since all major Internet cables in the US are located next to rail tracks (contrary to the popular belief that they follow electric transmission or phone lines). The telecom infrastructure has also become pervasive in this era of Big Data. As mentioned previously, smart meters are now being installed to monitor water and electricity consumption in real time. Moreover, data centers and other telecom infrastructure require large quantities of water for cooling. Similarly, they require enormous amounts of electricity. In fact, this is one of the reasons why Google and Facebook decided to install their main data centers in Oregon, thanks to its geographical proximity, relatively mild temperature, and ample source of water. Blum (2012) offers a fascinating account into the main telecom infrastructure: the Internet. 562 Environment and Planning B: Urban Analytics and City Science 44(3) a The pervasiveness of telecom also makes it an incredibly robust system (Newman et al., 2006). That being said, it can also be inherently fragile depending on how the data being collected or transmitted is stored. For instance, the telecom lines may be intact, but if the power system of a water treatment plant is down, no information can be transmitted. Its vulnerability may therefore be limited to the vulnerability of the other UIS for which it is being used. Finally, buildings are inherently integrated with telecom infrastructure as end points. Buildings The integration of buildings with all UIS is natural, and in fact, buildings are arguably the one common denominator in UIS planning. Put differently, all UIS service buildings, and whether it is the transport, water, utility, electricity, or telecom industry, all view buildings as demand points where the energy or resource being distributed is consumed, Considering the main objective of this work, buildings represent the only UIS that overlaps with all UIS. Buildings are therefore the end point for transportation (e.g. home and work); water for daily water needs; gas for space heating and cooking; electricity for all activities requiring electricity; and telecom for phone, television, and internet access. The integration of UIS should therefore enable an equally good or better service to buildings, while being more sustainable and resilient. The role of UIS integration At this point, we have seen how UIS are naturally highly interdependent. While having the various UIS independently planned by specific agencies may have been desirable in the 20th century (following a tree structure), since it was the first time they reached such as large scale and level of complexity, the current practices should change in the future, fostering more communication and coordination across departments as suggested by Minett (1975). Therefore, there exists much potential to plan them together, as a semilattice, so as to be better able to address the current environmental crisis for example and make UIS more resilient; e.g. to be better able to handle extreme weather events. In fact, the current, largely unplanned, integration between UIS has created critical dependencies, eventually hindering the resilience of UIS (Buldyrev et al., 2010; Fu et al., 2014; Khoury et al., 2015). Until now, I have used the term integration fairly loosely, but we also need to account for decentralization. Two UIS can be deeply integrated in a highly localized manner, for example, water pumping stations require significant amounts of energy only at a selected few locations, or they can be integrated in a much more distributed manner, for example, telecom lines are becoming heavily integrated with water, gas, and electricity with the emergence of smart meters. ‘Accordingly, we can think of an integration-decentralization (ID) matrix as illustrated in Figure 3. The x-axis spans from highly centralized to highly decentralized, and the y-axis spans from completely separated to highly integrated. I will use four examples to illustrate each quadrant of the matrix. Current paradigm First, in the bottom left quadrant, little integration is planned and UTS are centralized, which T would argue reflects the overwhelming majority of the current practice. A good example of Derrible 563 Integrated Localized Integrated Integration | Decentralization Centralized — Current Siloed Decentralized Paradigm Distribution Separated Figure 3. Integration-decentralization (ID) matrix this is the Chicago Tunnel And Reservoir Plan (TARP). Like most cities that are largely paved (thus having mostly impervious surfaces), Chicago suffers from flooding during heavy rains. This is accentuated by the fact that Chicago has a combined sewer system, and any surface runoff inevitably ends up in the sewer system. The system therefore fills up quickly and the ‘wastewater treatment plants cannot handle the sudden influx in wastewater. To remediate this problem, in the 1960s, the region decided to implement a deep tunnel system to dump wastewater during heavy rain events and pump it back to the surface for treatment once the rain stops and the plant can handle it. For more information about the Chicago TARP system, the reader is referred to Scalise and Fitzpatrick (2012). Notwithstanding the fact that the efficieney of the system is questionable, the location of these deep tunnels follows a typical tree structure. As shown in Figure 4(a), the tunnels are mostly located below rivers. This decision was made partially because overflowing untreated wastewater was simply dumped in the rivers. Another substantial reason, however, is because the water department owns all waterways, therefore avoiding rights-of-way (and administrative) conflicts. Put differently, although it might have made more sense to locate these tunnels below underdeveloped areas, such as highways, which could have significantly decreased construction costs, the decision was made to keep the entire system separated, A similar example exists with the Boston “Big Dig,” where elevated expressways were torn down and underground tunnels were built to handle the exact same traffic. Instead of thinking of a holistic solution to rid the city of elevated expressways by distributing the traffic across the network and increasing transit service (as was done in Seoul with the Cheonggyecheon river), a similar expressway was built underground. Siloed distribution Second, in the bottom right quadrant, UIS remains separated, but they start to be decentralized. This quadrant reflects most of the type of decentralization that is thought of at the moment, and there are many examples of this. A particularly fitted example for this quadrant is the current movement toward the installation and use of solar photovoltaic (PV) panels (Figure 4(b)). Solar PV panels are installed on roofs to produce electricity during the day. The electricity can be consumed onsite or fed back to the power grid. This completely decentralized way to produce electricity directly contrasts with traditional means to produce power from large thermal plants, ranging from coal, oil, and natural gas to nuclear power plants, The main argument toward a greater use of solar panels stems from the fact that the energy source is renewable, and at the same time, they may provide greater resilience to a system since the failure of one solar panel would have little impact on the grid as opposed to ee eee 564 Environment and Planning B: Urban Analytics and City Science 44(3) Derrible 565 u purely decentralizing single UIS is desirable, there may be a way to integrate UIS at the a) . same time. mam A @ ° Figure 4. Illustrations for the ID matrix. (a) Chicago TARP system, (b) solar panel on a house, (c) concept for SMART tunnel, and (4) rain garden. the failure of a large power plant that can bring an entire region to a standstill. In contrast, however, the power produced by standard thermal power plants is constant and reliable as ‘opposed to the power produce by solar panels. Judging the benefits of solar panels is out of the scope of this article. ‘We should note that similar efforts are being put into decentralizing other UIS that tend to be separated, For example, new technologies are coming out to treat water much more locally (both for water and wastewater treatment), Treating drinking water at the building level for instance would offer significant benefits since it may relieve the necessity for current water departments to keep water pressure in conduits above a certain minimum (around 140 kPa or 20 psi (Chin, 2014)) for fear of contamination. In addition, related to the theme of flooding, rainwater can be collected in rain barrels, which could potentially prevent a significant amount of runoffs. Effort is also being put into engineering small wetlands to treat wastewater. “Most of the siloed distribution cases stem from experts to improve UIS from their field of expertise, which partially explains why these cases lack integration with other UIS, and while Localized integration ‘Third, in the top left quadrant, and in direct opposition to the previous section, UIS starts to be integrated but only in localized areas. This practice reflects well what is generally understood by UIS integration, where effort is put into integrating UIS, but this is only done in a sporadic fashion. A pertinent example to illustrate this quadrant is the Stormwater Management and Road Tunnel (SMART) tunnel in Kuala Lumpur, Malaysia. Like most cities in South-East Asia, Kuala Lumpur has a rainy season (monsoon), during which heavy rain events are frequent, easily causing flash floods. In any day, the SMART tunnel carries mostly vehicles, with potential runoffs flowing in the lower section of the tunnel (Figure 4(c)). During flash floods, however, the tunnel is closed to traffic and it is filled with stormwater. Unlike the case of Chicago, Kuala Lumpur is able to separate its stormwater from its sanitary sewers, and the water that is dumped in the tunnel therefore does not contain any sanitary sewage. The tunnel itself therefore integrates transport UIS with water UIS, The integration, however, is heavily localized since only the tunnel itself gets flooded. The deep lake water cooling system in Toronto is another good example. The lake provides a source of naturally cool water (at about 4°C toward the bottom of the lake in the summer), which can be used in a heat exchanger to chill water, which in turn cools down a building. The warmed up water from the lake is then simply transported to a water treatment plant to be treated and distributed in the city as drinking water. In this case, the utility UIS is integrated with the water UIS. These two examples therefore highlight the premise of a semilattice structure since a right of way or a resource is shared between two UIS despite the fact that two separate agencies are involved. As we increasingly talk of “infrastructure ecology” and UIS integration, this type of scenario is likely going to inerease in the future and may at the same time improve the resilience of UIS (Khoury et al., 2015). Integrated Decentralization Finally, in the top right quadrant, UIS are both integrated and decentralized. In this case, UIS are not only integrated locally akin to the previous example, but they are integrated over a large area of a city. This is arguably the type of integration that is most desirable in the general push for more sustainable and resilient cities. Remaining in the water realm, a fitting example involves various green infrastructure strategies, including permeable pavement, rain gardens, and bioswales, As the name suggests, permeable pavement provides pores for water to seep through the ground as opposed to turning into surface runoffs. Rain gardens and bioswales are patches of plants with deep roots and large rocks (to create voids) located near roads, parking lots, or even by rain gutters. Unlike grass that can absorb little rain (about 2.5m), deep-rooted plants can absorb and retain water in their roots for relatively long period of times, therefore both preventing runoffs and relieving the requirement to water them for extended period of times in dryer seasons. These rain gardens are extremely low cost and can be implemented in any street. By being located in streets, this water UIS becomes integrated with transport UIS. At the same time, they are heavily decentralized since they can be installed all over a city. In fact, many cities have now adopted the green infrastructure route to partially solve their urban flooding problem si 566 Environment and Planning B: Urban Analytics and City Science 44(3) (Sedlak, 2014). Chicago has several projects of that nature, including the Pilsen Sustainability Street (City of Chicago, 2012), where Figure 4(4) was taken. Other examples include the use of solar panels and small water turbines to power streetlights and traffic signals, or the use of electric vehicles to store energy. In the end, adopting this strategy requires most often a significant change in the state of the practice for many cities, sometimes involving multiple agencies, therefore perhaps creating more of a semilattice structure in how UIS are planned, Overall, this ID matrix offers a means to compare and classify UIS plans. While I will not argue that all future UIS plans should be both integrated and decentralized, effort should be put into better integrating UIS with one another, thus following a semilattice structure, The example for Chicago gives a pertinent illustration. First off, instead of entering the sewer system, rainfall can either be collected in rain barrels (siloed distribution) or seep through the ground (ID). If these two strategies fail, tunnels can be built in areas that require lower construction costs and that are more easily reachable, such as below highways (localized integration), even though water departments do not own the land occupied by highways. ‘The general practice of UIS planning may therefore benefit significantly from adopting more of a semilattice structure. Conclusion ‘The main goal of this article was to study the parallel between the lessons from Alexander’s A City isnot a Tree applied to urban form with how UIS are currently planned in cities. The main argument was the current practice in UIS planning has evolved to have separate agencies plan the various UIS with no or little planned or coordinated integration among one another. While it might have been necessary in the 20th century, since it was the first time that UIS were built at such a large scale and level of complexity, adopting a semilattice structure, as defined by Alexander, may be preferable for more sustainable and resilient designs. In this article, I first recalled the lessons of Alexander, discussing the differences between a tree and a semilattice structure. According to Alexander, mass plans have tree structures, which are reflected by a physical separation between elements of a city. This contrasts directly with older neighborhoods that have experienced a much more organic growth, thus enabling elements of a city to “overlap.” Quite naturally, I argue that UIS planning currently follows a tree structure since separate agencies are responsible for the planning and design of their own UIS without having to coordinate with other agencies. Second, I recalled that the various UIS are actually deeply interdependent by nature. In particular, Table 1 listed some of the ways that each UIS depended on one another. Electricity, for instance, depends on transportation for the provision of raw materials. It is competing for underground space with the water, utility, and telecom UIS. It also requires much water for power generation, and it is increasingly linked with the telecom UIS with the deployment of smart meters. Buildings then represent the locations where electricity is consumed. We saw that streets tend to be significant since they host transportation, electricity, water, utility, and telecom UIS. The third section related to the role of UIS integration. Specifically, an ID matrix was introduced. The current paradigm tends to have little integration and be heavily centralized (e.g. Chicago TARP system). The siloed distribution quadrant includes projects that attempt decentralization that strictly stay within the strict disciplinary boundaries of a UIS (e.g. solar panels, rain barrels). The localized integration quadrant emphasizes on integration but remaining fairly centralized (¢.g. SMART tunnel in Kuala Lumpur). Finally, the ID Derrible 567 offers both high integration among UIS and high decentralization across an urban environment (e.g. rain gardens, permeable pavement). Overall, systematically thinking about ways to integrate and decentralize UIS may provide significant benefits in our quest to design more sustainable and resilient cities, and this can only start by having sound UIS planning practices that adopt a semilattice structure as originally defined by Alexander in 1965. While developing clear guidelines for engineers and urban planners to promote a sound integration and decentralization of UIS remains challenging, we may end with this quote from Minett (1975) “it is the links that create the system,” and focusing on how to better “link” the various UIS should be the focus of future work. Declaration of conflicting interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. Funding The author(s) received no financial support for the research, authorship, and/or publication of this article. References Ahmad N and Derrible S (2015) Evolution of public supply water withdrawal in the USA: A network approach. Journal of Industrial Ecology 19(2): 321-330. Alexander C (1965) A city is not a tree. Architectural Forum 122(1): 58-62. Bazilian M, Rogner H, Howells M, et al. (2011) Considering the energy, water and food nexus Towards an integrated modelling approach. Energy Policy 39(12): 7896-7906, Blum A (2012) Tubes: A Journey to the Center of the Internet. New York City: HarperCollins. Buldyrev SV, Parshani R, Paul G, et al. (2010) Catastrophic cascade of failures in interdependent networks. Nature 464(7291): 1025-1028 Byers EA, Hall JW and Amezaga JM (2014) Blectricity generation and cooling water use: UK pathways to 2050. Global Environmental Change 25: 16-30. Chin DA (2014) Water-Resources Engineering. Upper Saddle River, NJ: Pearson Education. City of Chicago (2012) City unveils ‘Greenest Street in America’ in Pilsen neighborhood. City of Chicago. Available at: http://www.cityofchicago.org/city/en/depts/edot/provdrs/eonservation_ outreachgreenprograms/news/2012/oct/edot_opens_the_pilsensustainablestreet.html (accessed 21 April 2016). Dawson R (2007) Re-engineering cities: A framework for adaptation to global change. Philosophical Transactions of the Royal Society of London A: Mathematical, Physical and Engineering Sciences 365(1861): 3085-3098. Derrible S and Ahmad N (2015) Network-based and binless frequency analyses. PLoS ONE 10(11) 0142108, Derrible S and Kennedy C (2011) Applications of graph theory and network science to transit network design. Transport Reviews 31(4): 495-519. Derrible S and Reeder M (2015) The cost of over-cooling commercial buildings in the United States. Energy and Buildings 108: 304-306. DHS (2015) Critical infrastructure security. Department of Homeland Security. Available at: http:// wwww.dhs.gov/topic/critical-infrastructure-security (accessed 21 April 2016). DNAInfo (2016) Where exactly is your neighborhood? DNAMfo. Available at: https:// visualizations dnainfo.com/nycneigh{ (accessed 21 April 2016). 568 Environment and Planning B: Urban Analytics and City Science 44(3) Duefias-Osorio L, Craig JI, Goodno BJ, et al. (2007) Interdependent response of networked systems. Journal of Infrastructure Systems 13(3): 185-194. Duefias-Osorio L and Vemuru SM (2009) Cascading failures in complex infrastructure systems. Risk Acceptance and Risk Communication 31(2): 157-167 Fiksel J (2003) Designing resilient, sustainable systems. Environmental Science and Technology 31(23): 5330-5339. Fotinos D (2003) Deep lake water cooling. In: Power Engineering Society general meeting, 13-17 July 2003. Vol. 4, pp.1-2300. IEEE. Fu G, Dawson R, Khoury M, et al. (2014) Interdependent networks: Vulnerability analysis and strategies to limit cascading failure. European Physical Journal B 87:148. Holling CS (1973) Resilience and stability of ecological systems. Annual Review of Ecology and Systematics 4: 1-23. HUD (2015) HUD History. US Department of Housing and Development. Available at: http://portal ‘hud .gov/hudportal/ HUD%sre=/about/hud_history (accessed 21 April 2016). IEA. (2014) Linking Heat and Electricity Systems: Co-generation and District Heating and Cooling Solutions for a Clean Energy Future. Paris, France: International Energy Agency. Available at: https: //www.iea.org/publications/freepublications/publication/LinkingHeatandElectricitySystems. paf (accessed 21 April 2016). Kennedy C, Cuddihy J and Engel-Yan J (2007) The changing metabolism of cities. Journal of Industrial Ecology W1Q): 43-59. Kenway SJ, Lant PA, Priestley A, et al. (2011) The connection between water and energy in cities: ‘A review. Water Science and Technology 63(9): 1983. Kermanshah A, Karduni A, Peiravian F, et al. (2014) Impact analysis of extreme events on flows in spatial networks. 2014 IEEE International Conference on Big Data, 27-30 October, Washington DC, pp. 29-34. Khoury M, Bullock S, Fu G, etal. (2015) Improving measures of topological robustness in networks of networks and suggestion of a novel way to counter both failure propagation and isolation. Infrastructure Complexity 2(1): 1-20. Kundzewicz ZW, Kanae S, Seneviratne SI, et al. (2013) Flood risk and climate change: Global and regional perspectives. Hydrological Sciences Journal $9(1): 1-28. Larsen ER and Arango S (2013) Thinking about the future: System dynamics and the process of clectricity deregulation. In: Qudrat-Ullah H (ed.) Energy Policy Modeling in the 2ist Century. Understanding Complex Systems. New York: Springer, pp. 17-30. Lee EI, Criddle CS, Bobel P, et al. (2013) Assessing the scale of resource recovery for centralized and ‘satelite wastewater treatment. Environmental Science and Technology 47(19): 10762-1070. Liu J, Mooney H, Hull V, et al. (2015) Systems integration for global sustainability. Science 347(6225). Makropoulos CK and Butler D (2010) Distributed water infrastructure for sustainable communities. Water Resources Management 24(11): 2795-2816. Mohareb B, Derrible § and Peiravian F (2015) Intersections of Jane Jacobs’ conditions for diversity ‘and low-carbon urban systems: A look at four global cities. J. Urban Plann, Dev., 10.1061) (ASCE)UP. 1943-5444.0000287, 05015004. Minett J (1975) If the city is not a tree, nor is it a system. Planning Outlook 16(1-2): 4-18. Mumford L (1961) The City in History: Its Origins, Its Transformations, and Its Prospects. New Yo Harcourt, Brace & World. ‘National Archives (2015) Records created or inherited by the Transport Ministries, and by related bodies, and by the London Passenger Transport Board. The National Archives. Available at: http:|/ discovery.nationalarchives.gov.uk/details/1/C210 (accessed 21 April 2016). Neuman M and Smith § (2010) City planning and infrastructure: Once and future partners. Journal of Planning History 91): 21-42. Newman ME, Barabisi AL and Watts DJ (2006) The Structure and Dynamics of Networks. Princeton, NJ: Princeton University Press. Pagani GA and Aiello M (2011) Towards decentralization: A topological investigation of the medium ‘and low voltage grids. /EEE Transactions on Smart Grid 2(3): 538-547. cr ee eee Derrible 569 Pahwa S, Youssef M, Schumm P, et al. (2013) Optimal intentional islanding to enhance the robustness of power grid networks. Physica A: Statistical Mechanics and its Applications 392(17): 3741-3754. Pandit A, Jeong H, Crittenden JC, et al. (2011) An infrastructure ecology approach for urban infrastructure sustainability and resiliency. In: Power Systems Conference and Exposition (PSCE), 2011 IEEE|PES, Phoenix, AZ, 20-23 March 2011, pp.l-2. IEEE, DOI: 10.1109) PSCE.2011.5772587 . Peiravian F, Kermanshah A and Derrible S (2014 (2015) Spatial data analysis of complex urban systems. Proceedings ~ 2014 IEEE International Conference on Big Data, IEEE Big Data 2014, 7004405, pp. 1-6. Perrings C (1998) Resilience in the dynamics of economy-environment systems. Environmental and Resource Economics 11(3-4): 503-520. Qin Y, Curmi E, Kopec GM, et al. (2015) China’s energy-water nexus ~ Assessment of the energy sector's compliance with the ‘3 Red Lines’ industrial water policy. Energy Policy 82: 131-143. Rezaie B and Rosen MA (2012) District heating and cooling: Review of technology and potential enhancements. In: (1) Green energy; (2) Special section from papers presented at the 2nd international energy 2030 conference, vol. 93, pp.2-10. Sage D, Sircar I, Dainty A, et al. (2015) Understanding and enhancing future infrastructure resiliency: ‘A socio-ecological approach. Disasters 39(3): 407-426 Scalise C and Fitzpatrick K (2012) Chicago deep tunnel design and construction. In; Structures ‘Congress 2012. pp.1485-1495. American Society of Civil Engineers. Sedlak DL (2014) Water 4.0: The Past, Present, and Future of the World's Most Vital Resource. New Haven: Yale University Press. Sovacool BK. and Sovacool KE (2009) Identifying future electricity-water tradeoffs in the United States. Energy Policy 31(1): 2763-2773. US DOT (2015) Department of Transportation (DOT). AllGov. Available at: http://www.allgov.com/

You might also like