SEARCH

SEARCH BY CITATION

Keywords:

  • power system modeling;
  • modeling;
  • decision support systems;
  • load flow analysis;
  • distributed power generation

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. 1. INTRODUCTION
  4. 2. BACKGROUND AND RELATED WORK
  5. 3. SYSTEM MODELING METHOD
  6. 4. IES
  7. 5. CASE STUDY: MASDAR CITY
  8. 6. CONCLUSION AND FUTURE WORK
  9. ACKNOWLEDGMENTS
  10. REFERENCES
  11. Biographies

City infrastructure systems have distinct functions but are not isolated from one another, with interactions existing between these systems. Modeling these systems requires a focus on the system functions and interdependencies. Most models focus on system failures rather than the unexpected effects of design decisions in these systems. This paper presents a functional and spatial modeling framework suited for the representation of city infrastructure systems. This framework comprises a systematic process for breaking down the system into fundamental components and defining the relations between the system components. In addition, the spatial feature of the framework facilitates the synthesis, analysis, and evaluation of infrastructures based on their geographical locations and spatial orientations. This system modeling approach is used to design an Integrated Energy System (IES) model in order to exhibit the features of this framework. The IES consists of standard energy system estimation techniques and tools such as MATPOWER for load flow analysis, and is also used to execute a city case study. As a result, the advantages of the functional and spatial framework for modeling city infrastructures are presented.

1. INTRODUCTION

  1. Top of page
  2. ABSTRACT
  3. 1. INTRODUCTION
  4. 2. BACKGROUND AND RELATED WORK
  5. 3. SYSTEM MODELING METHOD
  6. 4. IES
  7. 5. CASE STUDY: MASDAR CITY
  8. 6. CONCLUSION AND FUTURE WORK
  9. ACKNOWLEDGMENTS
  10. REFERENCES
  11. Biographies

Designing city infrastructure system models requires extensive visualization and estimation methods capable of adequately representing the interactions within and between the city systems. This interaction between two systems, i.e., interdependency, is “a bidirectional relationship between two infrastructures through which the state of each infrastructure influences or is correlated to the state of the other” [Rinaldi et al., 2001: 14]. System models typically focus on the processes within a system, overlooking the interdependencies that are critical in establishing system requirements, determining system behaviors, and evaluating system performances. City systems such as the energy, water, transportation, building, and waste systems are increasingly interdependent, and it is important to take these interdependencies into account while modeling city systems. As a result, system modeling methods that incorporate system interdependencies are required.

In addition, interdependencies are significant in system representation since the bidirectional requirements between systems determine the level of system integration, and consequently the probability of cascading failures [Carreras et al., 2007]. Critical infrastructure interdependency research tends to focus on these failures, but there is also a need to consider the unexpected effects of design decisions within each system and between systems in a system of systems.

This paper describes a system modeling framework suited to infrastructure systems. This modeling framework comprises a systematic method for breaking down and connecting system components as well as a spatial modeling framework for representing these system components. The systematic process of breaking down and connecting system components enables the analysis of the system in three dimensions: analysis of the system's inputs, analysis of the system's hierarchy, and analysis of the system's interdependencies. The modeling framework adopts the hierarchical system decomposition method presented in Alfaris et al. [2010], and, therefore, system models developed using this framework have a layered structure ranging from viewing the system as a whole to fundamental system parameters. In addition, the multidomain formulation approach [Alfaris et al., 2010] adopted in this modeling framework enables the integration of the system interdependencies. The spatial aspect of the modeling framework incorporates the geographical orientations and physical locations of system infrastructure facilities, thus further enhancing the modeling results.

Furthermore, the modeling framework is suited for use in the development of Decision Support Systems (DSSs), with the aim of informing administrative design decisions related to city infrastructures without having to explore advanced engineering processes. In order to demonstrate the advantages of system development using the functional and spatial system representation model, an Integrated Energy System (IES) model has been developed [Adepetu et al., 2012] and executed in the form of a case study.

The contributions of this paper are:

  • Development of a method for modeling city infrastructure systems that incorporates the functional and spatial features of the infrastructure systems in order to improve the system modeling process and results. The method builds on previous work on system decomposition [Alfaris et al., 2010], spatial topologies applied in Geographic Information System (GIS) environments, and functional layers discussed in Tolone et al. [2004] and Grogan and de Weck [2012].
  • Integration of the critical system interdependencies in the modeling framework since, typically, city systems do not operate in isolation. These interdependencies ensure that the impact of a system on the other systems, and vice versa, is taken into consideration during the modeling process. This is an application of critical infrastructure interdependency studies done by de Porcellinis et al. [2008] and Tolone et al. [2004].
  • Application of the system modeling framework to the development of an Integrated Energy System (IES) model. The advantages of the IES over conventional energy modeling frameworks include the application of an hourly load flow analysis, Distributed Generation (DG) modeling, integration of critical system interdependencies, and incorporation of spatial features in scenario analysis. The load flow analysis and distributed generation modeling are implemented in the IES using MATPOWER developed by Zimmerman et al. [2011].
  • Implementation and analysis of a sustainable city case study that shows the application of the IES model and illustrates the meaning and viability of the obtained results.

Section 'BACKGROUND AND RELATED WORK' examines similar models and studies on interdependency analysis. Section 'SYSTEM MODELING METHOD' gives a detailed explanation of the system development and modeling method. Section 'IES' gives a detailed description of the IES. Section 'CASE STUDY: MASDAR CITY' presents an application of the energy model. Section 'CONCLUSION AND FUTURE WORK' summarizes the system development model and highlights areas of future work.

2. BACKGROUND AND RELATED WORK

  1. Top of page
  2. ABSTRACT
  3. 1. INTRODUCTION
  4. 2. BACKGROUND AND RELATED WORK
  5. 3. SYSTEM MODELING METHOD
  6. 4. IES
  7. 5. CASE STUDY: MASDAR CITY
  8. 6. CONCLUSION AND FUTURE WORK
  9. ACKNOWLEDGMENTS
  10. REFERENCES
  11. Biographies

This paper is based on prior research work on modeling complex sustainable systems [Alfaris et al., 2010] and the application of this modeling approach in developing a city-scale energy system model [Adepetu et al., 2012]. Alfaris et al. [2010] provide a systematic multidomain approach that can be used to simultaneously break down a complex system and integrate the resulting subsystems. Adepetu et al. [2012] present the IES, an integrated energy model and briefly discusses other City.Net systems, i.e., water, waste, transportation, and building. The current energy model in this paper takes some steps forward from Adepetu et al. [2012] such as the conversion of the modeling approach from the annual scale to the diurnal scale, the introduction of a load flow analysis (using MATPOWER), etc.

Energy models and simulations that have objectives similar to either City.Net or the IES include Systems Advisor Model (SAM), Homer, Energis, SynCity, Urban Infrastructure Suite (UIS), Land Use Evolution and Impact Assessment Model (LEAM), and UrbanSim. The structures of these models and the contribution of this work to existing models are subsequently discussed.

SAM was developed by the National Renewable Energy Laboratory (NREL) and simulates different energy generation methods such as Photovoltaics (PV), Concentrated Solar Power (CSP), solar water heating, wind, and geothermal. SAM presents energy-related and extensive financial outputs based on the values of variables provided by the user but does not incorporate multiple energy source-load interactions. Homer (also developed by NREL) is similar to the SAM but executes multiple energy source-load optimizations. However, neither SAM nor Homer incorporate system interdependencies.

MetroQuest is an application targeted at stakeholders and enables its users to observe and understand the future effect of current policies. MetroQuest is based on the integration of health, sustainability, and air quality models, and it is applied as a tool for estimating greenhouse gas reduction, regional growth administration, and transportation-related development [MetroQuest, 2012].

EnerGis [Girardin et al., 2010] is a GIS that models energy system processes in urban areas. EnerGis is used for finding means to improve the efficiency of energy systems, and to advance the integration of renewable energy technologies as modes of distributed generation. In particular, EnerGis includes the heating and cooling requirements of geographical regions using GIS information.

Another model similar to the IES is SynCity [Keirstead et al., 2009], which is also used to model energy systems in urban areas. SynCity incorporates the city layout and socioeconomic factors such as population activity in the energy modeling process.

Urban Infrastructure Suite (UIS), developed by the US National Infrastructure Simulation and Analysis Center (NISAC) comprises seven integrated modules that model urban populations and infrastructures [NISAC, 2012a]. These seven modules represent transportation, energy, water, telecommunications, mobility, epidemiology, and finance. The Interdependent Energy Infrastructure Simulation System (IEISS) module [NISAC, 2012b] focuses on the energy system, modeling the energy-related interdependencies with nonlinear complexity. The objective of IEISS is to determine the service-providing capability of an electric grid and natural gas network.

The Land Use Evolution and Impact Assessment Model (LEAM) [LEAMgroup, 2012] employs a dynamic modeling approach, with autonomous submodels that represent the land use transformation being executed concurrently after calibration. LEAM computes the economic, environmental, and social impacts of these transformations in land use, and can therefore compare and evaluate different scenarios resulting from different land use policies.

UrbanSim, an open source decision support system targeted at Metropolitan Planning Organizations (MPOs), models land use, finances, transportation and the environment in urban areas as a result of selected policies and infrastructures [UrbanSim, 2012]. As a result, MPOs can compare and assess growth management policies for urban areas. The UrbanSim model is based on the interdependencies between different but vital infrastructures.

The above-mentioned models, which are state-of-the-art models albeit for different purposes, either do not incorporate system interdependencies or lack a structured spatial framework as described in this paper. Critical infrastructure studies that focus on system failures include [de Porcellinis et al., 2008] and [Tolone et al., 2004]. Porcellinis et al. show a method of modeling heterogeneous and interdependent critical infrastructures that accounts for possible failures in infrastructure with different levels of accuracy. Tolone et al. also focus on understanding cross-dependencies in infrastructure, thereby estimating acceleration and chain reaction of failures in infrastructures. This paper presents interdependencies with a focus on determining the effect of design decisions, rather than system failure that some of the aforementioned systems are designed to model.

Furthermore, a spatial modeling framework that enhances system design and could prove to be a powerful tool in system representation models is introduced in this paper. These two major aspects of the system model are further described in Section 'SYSTEM MODELING METHOD'.

3. SYSTEM MODELING METHOD

  1. Top of page
  2. ABSTRACT
  3. 1. INTRODUCTION
  4. 2. BACKGROUND AND RELATED WORK
  5. 3. SYSTEM MODELING METHOD
  6. 4. IES
  7. 5. CASE STUDY: MASDAR CITY
  8. 6. CONCLUSION AND FUTURE WORK
  9. ACKNOWLEDGMENTS
  10. REFERENCES
  11. Biographies

The functional and spatial system model comprises two research methods that complement each other. The first is a system functional representation that uses hierarchical decomposition and the multidomain formulation for modeling systems [Alfaris et al., 2010], comprising four stages: conceptualization, decomposition, formulation, and simulation. The second aspect of the research method is a spatial modeling framework that provides the infrastructure components with a geographical orientation, and informs the synthesis of components, analysis of behaviors, and evaluation of performances.

3.1. System Functional Representation

The four stages comprising the system functional representation process are described as follows.

3.1.1. Conceptualization

Conceptualization is the process of elaborating the fundamental ideas of the system. Conceptualization involves eliciting and specifying the requirements for the system to achieve its objective. One of the fundamental ideas of the system is the system process in Figure 1, specifying the function of each stage and the information flow between different stages in the developed system model.

image

Figure 1. System process.

Download figure to PowerPoint

3.1.2. Decomposition

Decomposition involves breaking down the system into components and logical phases, using a hierarchical approach and feeding the output from one stage as an input to the next stage. In general, decomposition determines the Form Parameters (FPs), Behavior Parameters (BPs) and Key Performance Indicators (KPIs), and these are the parameters at the synthesis, analysis, and evaluation stages, respectively. Decomposition with respect to system synthesis is displayed in Figure 2, logically separating the main structures in a system (nodes) from the links within and between these structures (edges). The decomposition process is hierarchical, breaking down a system one layer after another until the fundamental components of the system (represented by FPs) are defined.

image

Figure 2. Decomposition (synthesis). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

Figures 3, 4, and 5 show a sample decomposition process in progression. Figure 3 shows the decomposition of the wind farm with respect to the synthesis of the wind farm and how the fundamental FPs of the wind farm are defined. Figure 4 shows the second level of the system hierarchy with respect to the behavior analysis. The generation-consumption approach is employed since city infrastructure systems are either supplying or utilizing some resources, finances, etc. Based on the nature of city systems, four categories of generation-consumption behaviors were specified: resources, finances, emissions, and utilities/services. Looking downwards in the system hierarchy, Figure 5 shows the electricity generation behavior and how decomposition proceeds from resource generation to the different modes of electricity generation.

image

Figure 3. Decomposition of wind farm (synthesis). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

image

Figure 4. Decomposition of system behaviors. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

image

Figure 5. Decomposition of electricity generation. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

3.1.3. Formulation

Formulation comprises the identification of the parameter relations and the energy system's governing equations, i.e., relationships between the FPs, BPs and KPIs and the constraints involved. The formulation process is applied across every level of the system hierarchy as defined in the system decomposition and establishes the relationships between the different levels of the hierarchy. The formulation of wind electricity generation is seen in Figure 6, resulting in Eq. (1). This formulation is a continuation of the decomposition process in Figure 5.

image

Figure 6. Wind power generation analysis. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

3.1.4. Simulation

Simulation is the development of a software tool that models the system parameters, relations, and interdependencies. The BPs and KPIs are computed, and the results are presented as numbers and figures.

3.2. Spatial Modeling Framework

The spatial modeling framework, used to represent infrastructure system components, comprises nodes, edges, cells, and layers. The focus of this work is on city infrastructure systems, and, as a result, the geographical orientation of the system components contributes to the system behaviors and performances. This framework is similar to the framework used in Geographic Information System (GIS) applications [ESRI, 2012]. Figure 7 [Grogan and de Weck, 2012] shows a graphic description of the nodes, edges, cells, and layers. It is important to clarify that the layers in the system are only conceptual and functional, and do not correspond to physical layers. Conversely, the locations of the nodes and edges in the model represent the geographical locations of the represented system components. A cell represents a fundamental unit of space, and a combination of cells form a city layout.

image

Figure 7. City infrastructure spatial framework. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

Nodes represent points of interest in the infrastructure system while an edge is a connection between two nodes, carrying the “flow” between the nodes. For example, in the energy system, nodes include power stations, distribution substations, buses, and transformers while the edges are the power lines with different voltage levels that connect these nodes. It should also be pointed out that multiple nodes and multiple edges can occupy the same functional layer. The purpose of the functional layers is to clearly identify and group the functionalities of the nodes and edges in a system. Furthermore, the spatial modeling has the advantage of including the spatial parameters such as distance in the synthesis and analysis stages of the model, which makes the modeling process more comprehensive.

3.3. Model Structure

City.Net has a three-stage model structure [Adepetu et al., 2012] as seen in Figure 1. These stages are synthesis, analysis, and evaluation. These are the stages at which the resulting IES model can be used.

  1. Synthesis is the process of defining a custom system configuration for a scenario by defining the values of the FPs and the spatial properties of system components. For example, synthesizing a wind farm would involve defining the values of FPs such as number of wind turbines, turbine cut-in speed, turbine rated speed, turbine cut-out speed, etc. In addition, a spatial property of the plant such as its location would be defined in order to account for the distance of the wind farm to the various loads.
  2. Analysis is the stage of the defined model where the behavior of the system is obtained based on its defined configuration; i.e., the values of the BPs are obtained based on the defined values of the FPs.
  3. Evaluation involves rating the behaviors of the user-defined system using performance benchmarks, i.e., KPIs. In order words, the user-defined system is evaluated.

4. IES

  1. Top of page
  2. ABSTRACT
  3. 1. INTRODUCTION
  4. 2. BACKGROUND AND RELATED WORK
  5. 3. SYSTEM MODELING METHOD
  6. 4. IES
  7. 5. CASE STUDY: MASDAR CITY
  8. 6. CONCLUSION AND FUTURE WORK
  9. ACKNOWLEDGMENTS
  10. REFERENCES
  11. Biographies

The IES is defined according to the research approach specified in Section 'SYSTEM MODELING METHOD'. The FPs, BPs, and KPIs are outlined, and parameter relations are defined.

4.1. Conceptualization

The concept behind each aspect of the IES is detailed here. This includes subsystems in energy generation, transmission, and distribution.

4.1.1. Wind Farm

A wind farm consists of a number of wind turbines arranged for optimal energy generation. For wind farm arrangement, the distance between wind turbines on the same row should be about three to five turbine diameters and the distance between wind turbines on the same column should be about five to eight diameters [Goebel, 2010b].

4.1.2. PV

The PV station comprises several PV panels with the important factors affecting electricity generation being the total PV panel area, the panel efficiency at different temperatures and light intensities, and the hourly Direct Normal Irradiation (DNI) at the selected geographical location. The effect of temperature on PV power is important as temperature varies with time and location.

4.1.3. CSP

The different types of CSP stations include the parabolic trough, central tower, parabolic dish, and Fresnel mirrors. However, the parabolic trough type is the only CSP type currently synthesized in the IES.

4.1.4. Hydropower

The generation of energy from a hydropower station involves building the plant across a river, creating a head to yield potential energy and possibly harnessing kinetic energy from the river flow. The three major types of hydropower stations are impoundment, run-of-river, and pumped storage. The impoundment type taps into the potential energy while the run-of-river takes advantage of the river's kinetic energy [Tester et al., 2005].

4.1.5. Biomass

The biomass power station in the IES can be used in two modes: combustion or gasification of biomass. The combustion process uses a steam turbine for energy generation while the gasification process uses a combined cycle [Caputo et al., 2005]. As detailed in Caputo et al. [2005], the efficiency of the plant is dependent on the power capacity of the biomass station. Typically, higher power capacities have higher energy conversion efficiencies.

4.1.6. Natural Gas

The IES natural gas station follows the template applied by Spath and Mann [2000] and consists of a combined cycle. The combined cycle consists of a gas turbine and a steam turbine. The steam turbine capacity is typically about half the capacity of the gas turbine.

4.1.7. Transmission and Distribution

The IES utilizes load flow analysis algorithm for modeling power transmission and distribution. The user defines the electric grid configuration that comprises the bus, line, and transformer parameters. The load at each bus, which can be defined by the user based on a seasonal profile, is analyzed in conjunction with the seasonal variation of the generator capacities and the grid capacity. The load flow analysis is executed using MATPOWER [Zimmerman et al., 2011], a set of MATLAB files that use standard load flow analysis methods in analyzing power systems. Load flow analysis methods include the Newton-Raphson method and the Gauss-Seidel method.

The power grid structure and the load flow analysis approach applied in the IES ensure that distributed generation sources can be incorporated in the power grid. The load flow analysis is executed each hour of the year based on the variation of the capacity of the intermittent power sources (wind, PV, CSP) and the user-defined operating schedule of other power stations. As a result, the model captures the effects of the intermittences on meeting the load demand, the voltage balance of the power grid, and the effectiveness of distributed generation sources.

4.1.8. Revenue Generation

Revenue can be generated in the energy system by selling electricity to residential, commercial, and industrial consumers, which are represented in the other City.Net systems. Energy generated but not consumed locally by the load is injected to the external grid if the use case being modeled has a distribution system connected to the grid. Otherwise, excess energy is dumped.

4.1.9. Finance Consumption

Expenses in the energy system are generally represented in the form of initial capital expenditures and annual costs such as operational, maintenance, and fuel costs.

4.1.10. Emissions

There are generally two concepts of emissions: lifecycle emissions and actual emissions that occur during energy generation. The lifecycle emissions are used in estimating emissions from renewable energy methods such as PV and wind. These lifecycle emissions are the Greenhouse Gases (GHGs), usually represented in a CO2 equivalent, produced while manufacturing the equipment used in the power plant [Evans et al., 2009]. Typically, the lifecycle emissions are estimated per unit energy generated by a power plant, while those for the actual emissions (CO2, CO, CH4, and NOx) are estimated based on the amount of fuel used.

4.1.11. Resource Consumption

Resource consumption in the energy system is, for the most part, in the form of water use and land requirements:

  1. Water Consumption: In the energy system, water is used in the CSP, hydropower, natural gas, and biomass power plants. In the hydropower station, the water used is more of a case of water reserved and evaporated due to the water storage built in a hydropower station [Caputo et al., 2005].
  2. Land Requirements: Infrastructure such as power plants and substations require dedicated land space. This is the land space that is estimated in the IES and not the lifecycle land requirements. Lifecycle land requirements is a way of estimating the land footprint of energy generation technologies, ranging from the land used for manufacturing of the power station equipment to the disposal or recycling of this equipment after decommissioning the power station. This is the approach taken by Evans et al. [2009] in describing the land use as a sustainability indicator. The model in this paper focuses only on the land occupied by the power station during its years in operation. For infrastructure equipment, such as PV stations, CSP stations, and wind farms, the land required per unit power capacity is be estimated based on the spatial arrangement of PV panels, solar mirrors, and wind turbines, respectively.

4.2. Decomposition and Formulation

System parameters, parameter relations, and interdependencies of the IES are presented and discussed in this section. It is important to point out that certain assumptions have been made in order to make the model generally applicable and comprehensive at the same time. One broad assumption that has been applied to the modeling process is that of linear estimation; i.e., the behavior parameters such as the land occupied and costs vary linearly with the capacity of the power stations. According to the IES model, if a 5-MW plant has a capital cost of $10,000, then a 10-MW plant has a capital cost of $20,000, while this is not necessarily true in reality. However, the assumption might be reasonable taking into account the fact that the cost per MW ratio drops as plant capacity increases due to the economies of scale.

Another assumption is that of efficiency and operation hours in the plants such as the biomass power plant or natural gas plant. The efficiency typically increases (but at a decreasing rate) with the plant capacity increase and then hits a threshold. Moreover, the IES model assumes that the biomass or natural gas stations are either working at full capacity or not at all, but, in reality, they could work at any other capacity, e.g., 50% capacity or 60% capacity. Assumptions are also made in the case of the intermittent power sources such as solar and wind since the exact availability of these resources cannot be determined in advance.

The IES model does not apply a stochastic modeling approach. Therefore, these assumptions are necessary and aid the simplicity of the IES model. Sensitivity analysis for parameters such as the specific land use, specific emissions, specific water consumption, hourly wind speeds, and hourly solar irradiation can be varied in order to get a range of performance of modeled energy system scenarios. These seem to be the most fragile parameters for which it is necessary to perform extensive uncertainty analysis. In addition, the load can be used as the parameter of interest in a sensitivity analysis in order to estimate the impact of population growth in the city. As a result, the corresponding energy infrastructure development requirements can be determined with more precision.

The system FPs and BPs, and the relations between these parameters are as follows.

4.2.1. Wind Farm [Jahanbani-Ardakani et al., 2010]

The primary BP of the wind farm is the electricity generated E (MWh), and it is estimated based on the variation of the wind turbine power with wind speeds as seen in Eq. (1). The wind farm FPs are: wind speed v (m/s), duration of operation T (s), turbine cut-in speed v in (m/s), turbine cut-out speed v co (m/s), turbine rated speed vrated (m/s), turbine blade length r (m), turbine rated capacity Prated (MW), number of turbines in wind farm, N, and wind variation exponent m. The wind farm capacity Pstation (MW) is another BP that is determined based on the sum total of the wind turbine power ratings.

  • display math(1)
4.2.2. PV Station [Tester et al., 2005; NREL, 2012; Tamizhmani et al., 2003]

The electricity generated by the PV station E (MWh) is calculated based on the hourly DNI through the course of the year. One of the important factor s that affects the efficiency of the PV panel η is the module temperature Tmod (°C). Tmod is estimated hourly using the following FPs: ambient temperature Tamb (°C), hourly DNI (kWh/m2), and hourly wind speeds vwind (m/s):

  • display math(2)

In addition to Tmod, the energy generated per hour is calculated from the overall DC-to-AC derate factor df, PV panel length l (m), PV panel width w (m), PV panel efficiency, PV panel capacity Prated (kW), temperature coefficient of rated power, δP/T (%/°C), and number of PV panels N. df is used to estimate losses from panel rating errors, wiring, inverters, transformers, etc.

  • display math(3)
4.2.3. CSP Station [Tester et al., 2005; Goebel, 2010a]

The energy generation BP E (MWh) calculation of the CSP station is similar to that of the PV station but without temperature dependency. The FPs are: mirror aperture area A (m2), number of mirrors, N, plant solar-to-electric efficiency ηS − E, and hourly DNI (kWh/m2). ηS − E represents the plant efficiency from the solar field to the point of energy output (typically electricity).

  • display math(4)
4.2.4. Hydropower [Tester et al., 2005]

The energy generated E (MWh) by the hydropower station is calculated from the hydropower station rated capacity Pstation (MW) and the capacity factor CF, which represents how often the station is used at its full capacity. The other FPs are: the head h (m), inlet water speed vin (m/s), outlet water speed vout (m/s), volumetric flow rate Q (m3/s), plant efficiency η, capacity factor CF, water density ρ (kg/m3), and acceleration due to gravity g (m/s2):

  • display math(5)
  • display math(6)
4.2.5. Biomass [Caputo et al., 2005] and Natural Gas [Spath and Mann, 2000]

The same set of parameters are used in estimating the energy generated E by the biomass and natural gas power stations. The difference is the type of fuel used, i.e., biomass and natural gas as suggested by the power station names. The parameters used for estimating energy generation are: steam turbine capacity Pst (MW), gas turbine capacity Pgt (MW), and the hours of operation OH (h). In reality, the capacity of the turbines and the capacity of the power block are matched. Both steam and gas turbines are used in combined cycles while only one of either type of turbine is used in a single cycle.

  • display math(7)

Furthermore, the fuel flow rate, i.e., the mass of fuel consumed by the power station M (tonnes/year) is determined based on the generated energy, the fuel heating value HV (kJ/kg), annual operating full load hours OH (h), and plant energy-conversion efficiency at rated power ηp:

  • display math(8)
4.2.6. Finances [NREL, 2008; Tester et al., 2005]

In the IES, revenue R ($/year) is generated from the sale of energy. The energy sold is determined based on the energy actually used by consumersEC(MWh/year) and not the generated energy as there are some losses in the grid. The other FP included in the revenue estimation is the unit selling price of electricity SP (/MWh).

  • display math(9)

On the other hand, costs are estimated based on unit capital cost CunitCap ($/MW), unit operational and maintenance (O&M) costCunitO &M($/MW), unit fuel cost CunitF ($/kg), plant capacity P (MW), mass of fuel consumed annually M (kg), capital costCCap($), annual O&M cost CO &M ($/year), and the annual fuel costCF($/year). CF applies to power stations that use fuels such as the biomass and the natural gas.

  • display math(10)
  • display math(11)
  • display math(12)

These costs by themselves are not particularly useful for deducing the actual cost of generating energy, and this is where the levelized cost LEC ($/kWh) is applicable. The additional parameters used in estimating the LEC are: expected power station lifetime T (years), net cost in year t,Ct($/yr), energy generated in year t, Et (kWh), and discount rate d.

  • display math(13)

where t is the year andCt = 0 is the same as CCap.

4.2.7. Emissions

Emissions are estimated linearly from the energy generated by each power station, Eplant (kWh), and the fuel used per year, Mfuel (kg) (if applicable). The key FPs are the mass of CO2 per unit energy, CO2/kWh (kg/kWh), and the emissions per unit mass of fuel GHG/kg. As a result, the lifecycle emissions GHGLC (kg) and the actual emissions during generation, GHGGen (kg), can be obtained:

  • display math(14)
  • display math(15)
4.2.8. Resource Consumption

The resource consumption BPs in the IES are the land occupied by power plants Lplant (sq-km) and water consumed by power plants Wplant (tonnes). These BPs are linearly estimated from the energy generated per year in the power plant, Eplant (kWh), water used per unit energy in the plant, WperkWh (tonnes/kWh), land required per unit kW in plant, LperkW (sq-km/kW), and infrastructure capacity Pplant (kW):

  • display math(16)
  • display math(17)
4.2.9. Transmission and Distribution [Zimmerman et al., 2011; Grainger and Stevenson, 1994]

Equation (18) is the Gauss-Seidel equation that finds a balance in the values of the voltage at each bus in a power grid. The FPs used are: the net load Pk + jQk (VA) at each bus, bus voltage Vk (V), and number of buses, N.

The admittance matrix Y− 1) is obtained from the impedances between buses. It is a product of the feeder cable impedance per unit distance and the length of the cable between buses. Y is calculated by the MATPOWER module.

  • display math(18)

where k is the number of the bus in consideration.

4.2.10. Energy-Related Interdependencies

Table I shows a summary of energy-related interdependencies in other systems, i.e., energy generation and energy consumption in other systems. These relations help in defining the role the energy system plays in other city infrastructure systems. This paper focuses on the IES alone and the interdependencies only show the first-order impacts of the system requirements. However, the nth-order impact, where n is the number of system BPs, of these interdependencies can be visualized in an integrated city model that includes parameters from all five City.Net systems. These nth-order impacts can be measured using multidomain matrices as defined in Alfaris et al. [2010]. This is an area for future research.

Table I. Energy Generation and Consumption in Other City.Net Systems
Image

Figure 8 provides an overview of the City.Net system from the energy system perspective. Figure 8 shows energy-related dependencies in the city infrastructure system following the hierarchy used in the decomposition template seen in Figure 4. Generation and Consumption refer to the generation and consumption of resources, finances, and emissions within the energy system, hence, the “G” and “C” tags used in other parts of Figure 8.

image

Figure 8. Energy system perspective of City.Net. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

4.2.11. KPIs

At the evaluation stage in IES, KPIs are used to estimate and rate the performance of the energy system. These KPIs are combinations of environmental, social, and financial indicators and they follow the equations in [Adepetu et al., 2012]. They are listed as follows:

  1. Renewable Energy Fraction (REF) [Economist Intelligence Unit and Siemens, 2009]
  2. Energy Cost Indicator (ECI) [Afgan and Carvalho, 2004]
  3. Capital Cost Indicator (CCI) [Afgan and Carvalho, 2004]
  4. Energy Consumption per Head (ECH) [Economist Intelligence Unit and Siemens, 2009]
  5. Energy Intensity (EI) [Economist Intelligence Unit and Siemens, 2009]
  6. CO2Emissions per Head [Economist Intelligence Unit and Siemens, 2009]
  7. CO2Savings
  8. Area per MW [Afgan and Carvalho, 2004].
4.2.12. IES Layers

The nodes and edges in the energy system are classified according to their functions in the energy system. The layers are not strictly ordered. For example, transmission substations on layer 2 can be eliminated in the cases of distributed generation. The classifications are listed in Table II.

Table II. IES Layers
Image

4.3. Simulation

Simulation of the IES comprises a Graphic User Interface (GUI), Microsoft Excel spreadsheets, a MATLAB Application Programming Interface (API), and MATLAB *.m files at the simulation stage. The synthesis process is executed with the GUI and spreadsheets, while analysis and evaluation is executed on MATLAB. The diurnal simulation time approach is applied in the IES. This way, the variations in the solar irradiation and wind speeds and the effects of these intermittences on the power grid are adequately incorporated in the IES modeling process. Schedules are also applied to the biomass and natural gas power stations where they are either in operation or not.

5. CASE STUDY: MASDAR CITY

  1. Top of page
  2. ABSTRACT
  3. 1. INTRODUCTION
  4. 2. BACKGROUND AND RELATED WORK
  5. 3. SYSTEM MODELING METHOD
  6. 4. IES
  7. 5. CASE STUDY: MASDAR CITY
  8. 6. CONCLUSION AND FUTURE WORK
  9. ACKNOWLEDGMENTS
  10. REFERENCES
  11. Biographies

Masdar City is located in the outskirts of Abu Dhabi, United Arab Emirates (UAE) with coordinates 24° 26’ 2.55" N and 54° 36' 14.44" E. Masdar City is still under construction and aims to be the world's first sustainable and carbon neutral city. The city is home to the Masdar Institute of Science and Technology and will house 50,000 people [Carvalho, 2009]. This case study illustrates a simplified application of the IES.

5.1. Synthesis

The energy infrastructure simulated are a wind farm, biomass power station, PV station, and alternating current (AC) distribution system.

5.1.1. Wind Farm

The wind farm has a capacity of 11 MW with four GE 2.75–103 [General Electric Power and Water, 2011] wind turbines. The hourly wind data for the Masdar City area was not available, but an hourly dataset with similar aggregate wind speeds was used.

5.1.2. PV Station

The PV station has a capacity of 10 MW and comprises solely Suntech STP280-24/Vd [Suntech, 2012] polycrystalline PV modules. The PV modules have a Standard Test Condition (STC) efficiency of 14.4%. Abu Dhabi has an annual DNI of about 2000 kWh/m2 and is therefore a good location for a PV station (the hourly solar radiation data was not available but a dataset with a total DNI of about 2000 kWh/m2 was used). In addition, the temperature variation in the Masdar City area was included in the synthesis of the PV station.

5.1.3. Biomass Station

The biomass power station has a capacity of 47 MW with a gas turbine of 30 MW and a steam turbine of 17 MW. A power plant of this output capacity and configuration is expected to have an efficiency of 45% [Caputo et al., 2005]. Also, a schedule is used to determine when the biomass power station is in operation or offline.

5.1.4. Distribution System

An AC distribution system is simulated with a mesh network as seen in Figure 9. Five buses are configured in the distribution system: one slack bus (Bus 1), two load buses (Bus 2 and Bus 4), and two voltage control buses (Bus 3 and Bus 5). The loads at the various buses use the IEEE RTS load model but have different peak values. The peak loads at the various buses used in the simulation are as follows: 10 MW at Bus 2, 55 MW at Bus 3, 8 MW at Bus 4, and 9 MW at Bus 5. By inspection, it can be observed that the generators cannot meet the above-mentioned loads. Therefore, the city is dependent on the utility to meet its energy demands. Figure 10 shows the connections between the generators and the buses in different layers.

image

Figure 9. Distribution system (layer 4).

Download figure to PowerPoint

image

Figure 10. Masdar City energy system layers. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

Since the Masdar City case study is a small city application of the IES model, there is no transmission layer as seen in Figure 10. That is, there are no transmission lines, and the city's electrical grid connects directly to the distribution level of the national grid. In case studies with larger cities, the transmission layer could be included depending on the grid structure.

5.2. Analysis and Evaluation

The analysis and evaluation results are shown in Tables III-VII. The results show a REF of 64.7%, resulting in CO2 emission savings of 66,609 tonnes/year. According to the Siemens European Green City Index [Economist Intelligence Unit and Siemens, 2009] compiled in 2009, the CO2 emissions per head across Europe ranged from 2.19 tonnes/year in Oslo to 9.72 tonnes/year in Dublin. Therefore, CO2 savings per head of 1.33 tonnes/year obtained from the simulation represents a significant contribution to the reduction of GHG emissions. This result is, however, dependent on the assumption that the load profile used in the simulation would be sufficient to represent the energy demand of 50,000 people in Masdar City.

Table III. Wind Farm FPs and BPs
Image
Table IV. PV Station FPs and BPs
Image
Table V. Biomass Station FPs and BPs
Image
Table VI. Distribution System FPs and BPs
Image
Table VII. Masdar City KPIs
Image

The distribution system is also seen to be stable with a minimum voltage of 0.9846 p.u. and a maximum voltage of 1 p.u. at the voltage control buses. The total loss in the distribution system is 12.37 GWh/year and this is about 3% of the energy generated within the city. Figure 11 shows the energy generated by the wind farm and PV station through the course of the year. The peaks and troughs of the wind and PV stations can be visualized. However, the load flow analysis shows that the grid maintains a stable voltage with a voltage range of 0.98– 0.99 p.u. at Bus 2 and a stable voltage of 0.99 p.u. at Bus 4. As a result, the intermittencies in the power supply of the wind and PV stations can be adequately managed in the modeled power system configuration.

image

Figure 11. Hourly energy generation. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

It should be pointed out that the city is dependent on the utility for meeting the local energy demands as specified in the simulation. As seen in Table VI, 215.83 GWh is obtained from the utility grid and only 3.04 GWh is supplied in return to the utility grid annually. In order for the city to be independent of the utility grid, the capacities of the DG implemented in the case study would have to be increased and storage would be required to balance intermittences in wind and PV. Furthermore, for larger cities and regions, the DG sources would have to be of higher capacities in order to meet the scaled-up energy demands.

One major setback is the absence of energy generation and consumption metrics in other City.Net systems due to this stand-alone execution of the IES. As a result, the impact of other systems on the energy system cannot be known. This reveals the importance of having complete system interdependency estimations.

5.3. Validation of Results

Estimations of similar proportions in the Masdar City case study were carried out using Homer and SAM, which are energy simulation applications. The results for the energy generated and levelized costs obtained are compared in Table VIII. Homer and SAM are used to validate the energy generation results obtained in the IES simulation. A significant number of other parameters such as LEC, water use, bus voltages, power system losses, and emissions are estimated based on the generated energy. As a result, it was important to verify that the energy generation values were indeed accurate. In addition, the LEC values are validated as well as there are typically different methods for estimating levelized costs.

Table VIII. Comparison of Simulation Results
Image

The results obtained are similar, and this aids in validating some of the results obtained in the IES. It is important to point out that the input parameters could not be exact due to the different approaches taken by the different energy models, and this partly explains the different values obtained. Also, the simulation for biomass power in the SAM had many high-detail parameters and, as a result, was not suitable for comparison with the IES.

Additional validation of the results can be achieved by performing sensitivity analysis and uncertainty analysis as a part of the city modeling process. As mentioned previously in the paper, assumptions were made with respect to some parameter values and relations. Sensitivity and uncertainty analyses would establish a range of results obtainable based on the assumptions made in the model, thus improving the credibility of the IES model and the ability of the IES model to support decision making. Validation by sensitivity analysis and uncertainty analysis, and their efficient incorporation in the modeling process, is an area of consideration for future research work. As a promising direction, this validation could include the application of the one-factor-at-a-time (OAT) analysis approach in order to evaluate the impact of singular parameters of interest. It could also include execution of the analysis with a focus on combinations of closely linked parameters of interest.

The primary limitation of any conclusions drawn from the Masdar City case study is that the energy demand used in the simulation does not represent the actual energy consumption within Masdar City at the time when the city is fully completed. The city is still under construction, and, as a result, the actual energy demand patterns of the city at full operation cannot be fully determined. This extends to the analysis and sharing of any economic issues and conclusions based on this study. However, the main purpose of the case study is to show how City.Net IES can be used to estimate energy system interactions based on the expected energy demands.

6. CONCLUSION AND FUTURE WORK

  1. Top of page
  2. ABSTRACT
  3. 1. INTRODUCTION
  4. 2. BACKGROUND AND RELATED WORK
  5. 3. SYSTEM MODELING METHOD
  6. 4. IES
  7. 5. CASE STUDY: MASDAR CITY
  8. 6. CONCLUSION AND FUTURE WORK
  9. ACKNOWLEDGMENTS
  10. REFERENCES
  11. Biographies

A functional and spatial modeling framework suited for the representation of city infrastructure systems is described in this paper. The functional aspect of the framework is based on a hierarchical decomposition and multidomain formulation approach for the modeling of complex sustainable systems. The spatial aspect of the framework comprises a spatial representation structure that enables the synthesis, analysis, and evaluation of infrastructures based on their geographical locations and spatial orientations. The properties of this framework are exhibited in the development of the IES and a Masdar City case study. The IES consists of standard energy generation technologies and utilizes MATPOWER as a load flow analysis tool. As a result, the IES incorporates the simulation of active generation networks, which is currently an area of interest in the power systems field, particularly in the grid integration of renewable energy technologies.

One area of future interest is the development of the City.Net model into an executable DSS. As mentioned previously, the functional and spatial model can be used to develop a system model for a DSS, and Georgilakis [2006] provides benchmarks for state-of-the-art energy DSS. Another area of future work is the integration of the functional and spatial system model with GIS applications. This is made feasible by the spatial framework that makes up part of the model.

Concisely, the modeling framework introduced in this paper is a modeling approach useful for modeling city infrastructure systems since it systematically defines the fundamental system components, accounts for interdependencies, and utilizes a spatial system representation.

ACKNOWLEDGMENTS

  1. Top of page
  2. ABSTRACT
  3. 1. INTRODUCTION
  4. 2. BACKGROUND AND RELATED WORK
  5. 3. SYSTEM MODELING METHOD
  6. 4. IES
  7. 5. CASE STUDY: MASDAR CITY
  8. 6. CONCLUSION AND FUTURE WORK
  9. ACKNOWLEDGMENTS
  10. REFERENCES
  11. Biographies

This work was supported by funding from a MIT-Masdar Institute of Science and Technology collaborative research grant, Project Code 400030.

REFERENCES

  1. Top of page
  2. ABSTRACT
  3. 1. INTRODUCTION
  4. 2. BACKGROUND AND RELATED WORK
  5. 3. SYSTEM MODELING METHOD
  6. 4. IES
  7. 5. CASE STUDY: MASDAR CITY
  8. 6. CONCLUSION AND FUTURE WORK
  9. ACKNOWLEDGMENTS
  10. REFERENCES
  11. Biographies

Biographies

  1. Top of page
  2. ABSTRACT
  3. 1. INTRODUCTION
  4. 2. BACKGROUND AND RELATED WORK
  5. 3. SYSTEM MODELING METHOD
  6. 4. IES
  7. 5. CASE STUDY: MASDAR CITY
  8. 6. CONCLUSION AND FUTURE WORK
  9. ACKNOWLEDGMENTS
  10. REFERENCES
  11. Biographies
  • Image of creator

    Adedamola Adepetu is a Ph.D. student in the Cheriton School of Computer Science at the University of Waterloo, Canada. His research interests include energy systems, system modeling, and crowd sourcing. His current work aims to understand seasonality in electricity and gas loads using machine learning methods. He holds a master's degree in Computing and Information Science from Masdar Institute of Science and Technology, UAE, and a bachelor's degree in Electronic and Electrical Engineering from ObafemiAwolowo University, Ile-Ife, Nigeria.

  • Image of creator

    Paul Grogan is a Ph.D. candidate in the MIT Engineering Systems Division. His research interests include information system design, systems analysis, and simulation. His dissertation seeks to develop interactive simulation games for infrastructure system-of-systems design with an emphasis on supporting collaboration among independent decision-makers. He holds a master's degree in Aeronautics and Astronautics from MIT and a bachelor's degree in Engineering Mechanics and Astronautics from the University of Wisconsin, Madison.

  • Image of creator

    Anas Alfaris is currently an Assistant Professor at King Abdulaziz City for Science and Technology (KACST) as well as a visiting Assistant Professor at MIT. He is currently the codirector of the Center for Complex Engineering Systems at KACST and MIT. Prior to that, he was a Research Scientist in the Engineering Systems Division (ESD) at MIT. There he has led several research initiatives, advised several graduate students, and taught several courses focusing on Multidisciplinary System Design Optimization (MSDO) as well as Computer Aided Design (CAD). Dr. Alfaris received his training in several disciplines including civil architecture, engineering, and computer science. He started his career by earning a bachelor degree in Architecture and Building Engineering from King Saud University in Riyadh, Saudi Arabia. He then received a Master in Building Technology followed by a Master of Science in Architectural Studies with a focus on computational design systems, both from the University of Pennsylvania, Philadelphia. He subsequently received a Master of Science in Computation for Design and Optimization from the Center for Computational Engineering while completing his Ph.D. in Design Computation, both from MIT. His research experience spans several fields including Systems Architecture & Engineering, Generative Synthesis Systems, Integrated Modeling and Simulation, Multidisciplinary Analysis and Optimization as well as Decision Support Systems. His current research focuses on the development of Computational Design Systems for the design of Complex Engineered Systems. Furthermore, He has been part of several multidisciplinary research teams including the Design Lab and the Smart Cities Group at the Media Lab, MIT, Massachusetts and the Strategic Engineering Group at the Engineering Systems Division, MIT, Massachusetts.

  • Image of creator

    Davor Svetinovic is an assistant professor at Masdar Institute of Science and Technology, UAE, and a research affiliate at MIT. His current research interests include: strategic requirements engineering, systems architecture with emphasis on software and smart grids, and sustainable development from the systems security perspective. Previously, he worked at Lero—the Irish Software Engineering Center, Limerick, Ireland, and Vienna University of Technology, Austria. He received his Ph.D. and MMath degrees in Computer Science from University of Waterloo, Canada.

  • Image of creator

    Olivier L. de Weck focuses on how complex man-made systems such as aircraft, spacecraft, automobiles, printers, and critical infrastructures are designed and how they evolve over time. His main emphasis is on strategic properties that have the potential to maximize lifecycle value (aka the “lities”). Since 2001 his group has developed novel quantitative methods and tools that explicitly consider manufacturability, flexibility, robustness, and sustainability among other characteristics. Professor de Weck's teaching emphasizes excellence, innovation, and bridging of theory and practice. He is a Fellow of INCOSE and an Associate Fellow of AIAA. He serves as Associate Editor for the Journal of Spacecraft and Rockets and the Journal of Mechanical Design. He won six best paper awards since 2004, including the 2008 and 2011 best paper awards from the journal Systems Engineering. He won the 2006 Frank E. Perkins Award for Excellence in Graduate Advising, the 2010 Marion MacDonald Award for Excellence in Mentoring and Advising, and a 2012 AIAA Teaching Award. Since early 2011 he serves as Executive Director of the new MIT Production in the Innovation Economy (PIE) initiative. He has authored two books and about 200 papers.