ENERGY HUBS FOR THE FUTURE by M. Geidl, G. Koeppel, P. Favre-Perrod, B. Klöckl, G. Andersson, K. Fröhlich, Power Systems and High Voltage Laboratories, ETH Zurich, 8092 Zurich, Switzerland published in IEEE Power & Energy Magazine, 5(1):24–30, 2007 Important notice Copyright and all rights in this work are retained by the authors. This material may not be reposted without the explicit permission of the authors. ©2007 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from IEEE.

24 IEEE power & energy magazine january/february 20071540-7977/07/$25.00©2007 IEEE

©A

RT

VIL

LE, L

LC.,

PH

OT

OD

ISC

january/february 2007 IEEE power & energy magazine 25

MMOST OF TODAY’S ENERGY INFRASTRUCTURES EVOLVED DURING THEsecond half of the twentieth century, and it is questionable if they meet the requirements oftomorrow. Besides congested transmission systems, many facilities are approaching the endof their prospected lifetime. In addition, other issues such as the continuously growingdemand for energy, the dependency on limited fossil energy resources, the restructuring ofpower industries, and the general aim of utilizing more sustainable and environmentallyfriendly energy sources raise the question of whether piecewise changes of the existing sys-tems are sufficient to cope with all these challenges.

Various scientific studies have investigated future scenarios based on boundary condi-tions given by today’s structures, such as standardized electric voltage and gas pressure lev-els. Although these studies provide important insights, they often result in solutions thatcomply with the existing systems; possibly interesting and more long-term oriented solu-tions are hidden, as they lie beyond system-given boundaries. In contrast to these studies, aproject named “Vision of Future Energy Networks” was initiated at ETH Zurich togetherwith partners (see Table 1), which aims at a greenfield approach for future power systems.Restrictions given by the existing systems are basically neglected in order to determine realoptima. The consideration of multiple energy carriers, not only electricity, represents one ofthe key characteristics of this project. There is a belief that synergies among various forms of

energy represent a great opportunity for system improve-ments. Besides the possibilities of modern information tech-nology, state of the art as well as emerging and loomingenergy technologies, e.g., fuel cells, are taken into account.The time horizon for implementation is set to 30–50 yearsfrom now. Thus, the basic question to be answered is: “Howshould energy systems look in 30–50 years, and what can beexpected from them?”

Under these conditions, two key approaches are reasonable:transformation, conversion, and storage of various forms ofenergy in centralized units called energy hubs and combinedtransportation of different energy carriers over longer distancesin single transmission devices called energy interconnectors.

The project team soon realized that only a few established tools were available for the inte-grated analysis of multiple energy carrier systems, thus they focused in a first phase on develop-ing a modeling and analysis framework. In the second phase, which recently started, optimalsystem structures and operation strategies are determined and compared with conventional infra-structures using the developed tools. The result of this phase is the greenfield approach. Thefinal phase of the project is dedicated to identifying transition paths and bridging systems lead-ing from today’s systems to the identified optimal structures. Figure 1 outlines this process.

In the remaining part of this article, the key approaches, some developments, and firstresults of the project “Vision of Future Energy Networks” will be presented.

Combining Energy InfrastructuresIndustrial, commercial, and residential consumers requirevarious forms of energy services provided by different infra-structures. In the industrialized part of the world, coal, petro-leum products, biomass, and grid-bound energy carriers suchas electricity, natural gas, and district heating/cooling are

typically used. So far, the different infrastructures are con-sidered and operated almost independently. Combining thesystems can result in a number of benefits. Synergy effectsamong various energy carriers can be achieved by takingadvantage of their specific virtues. Electricity, for example,can be transmitted over long distances with comparably lowlosses; chemical energy carriers such as natural gas can bestored employing relatively simple and cheap technologies.With so-called line packing techniques, compressible fluidscan be stored in pipeline networks, even if there are no dedi-cated storage devices installed.

Combining the infrastructures means to couple them,thereby enabling exchange of power among them. Couplingsare established by converter devices that transform powerinto other forms. The question to be answered is, of course,

where to put which devices and how to oper-ate them. Answering this question is essen-tial for the system layout and therefore oneof the central issues in the project. There-fore, models and methods have been devel-oped to find the optimal coupling and powerexchange among multiple energy carriersbased on various criteria such as cost, emis-sions, energy efficiency, availability, securi-ty, and other parameters.

The Energy Hub ConceptThe key approach in the “Vision of FutureEnergy Networks” project is the so-calledenergy hub. An energy hub is considered aunit where multiple energy carriers can beconverted, conditioned, and stored. It repre-sents an interface between different energyinfrastructures and/or loads. Energy hubsconsume power at their input ports connect-ed to, e.g., electricity and natural gas infra-structures, and provide certain requiredenergy services such as electricity, heating,cooling, and compressed air at the output

ports. Within the hub, energy is con-verted and conditioned using, e.g.,combined heat and power technology,transformers, power-electronicdevices, compressors, heat exchang-ers, and other equipment. Real facili-ties that can be considered as energyhubs are, for example, industrialplants (steel works, paper mills), bigbuilding complexes (airports, hospi-tals, shopping malls), rural and urbandistricts, and small isolated systems(trains, ships, aircrafts). Figure 2shows an example of an energy hub.

The components within the hubmay establish redundant connections

figure 1. Transition from today’s system to the greenfield approach viabridging systems.

26 IEEE power & energy magazine january/february 2007

Time

OverallPerformance

Today'sSystem

BridgingSystems

Backcasting

Forecasting

2000 2050

"Optimal"

"Sub-optimal"

GreenfieldApproach

figure 2. Example of an energy hub that contains a transformer, a microturbine, aheat exchanger, a furnace, an absorption chiller, a battery, and a hot water storage.

Electricity

Natural Gas

Wood Chips

District Heat

Electricity

Heating

Cooling

Energy Hub

InputsOutputs(Loads)

table 1. “Vision of Future Energy Networks”ETH Zurich

Industry Academic AuthorityABB RWTH SwissAreva Aachen Federal

Office ofSiemens TU Delft EnergyPartners in the “Vision of Future Energy Networks” Project

january/february 2007 IEEE power & energy magazine

between inputs and outputs. For example, the electricity loadconnected to the hub in Figure 2 can be met by consuming allpower directly from the electricity grid or generating part orall of the required electricity from natural gas. This redundan-cy in supply results in two important benefits, which can beachieved using energy hubs. First, reliability of supply can beincreased from the load’s perspective because it is no longerfully dependent on a single network. Alternatively, reliabilityof the individual infrastructures could be reduced (e.g., byreducing maintenance) while availability for the load remainshigh. Second, the additional degree of freedom enables opti-mization of the supply of the hub. Energy carriers offered atthe hub’s input can be characterized based on their cost, relat-ed emissions, availability, and other criteria; the inputs canthen be optimally dispatched based on these quantities. Inaddition, utilizing energy storage represents an opportunity forincreasing the overall system performance, therefore, storageis already taken into account in the planning phase. Especiallywhen energy sources with intermittent primary energy (e.g.,wind, solar) are considered, storage becomes important sinceit enables affecting the corresponding power flows. Compen-sation of fluctuating power flows ispossibly the most evident applicationof energy storage technology. How-ever, investigations have shown thatstorage can be utilized in such a waythat it positively affects all of theaforementioned criteria, especiallywhen considering a liberalized mar-ket environment.

The InterconnectorConceptIntegrating different energy carriersis also possible in terms of transmis-sion. In the “Vision of Future Ener-gy Networks” project, a devicenamed energy interconnector is pro-posed that enables integrated trans-portation of electrical, chemical, andthermal energy in one undergrounddevice. So far, the most promisinglayout seems to be a hollow electri-cal conductor carrying a gaseousmedium inside (see Figure 3).

The basic motivation for combined transmission is thepossibility of efficiency improvement due to waste heatrecovery. The heat losses generated in the electrical con-ductor are partially stored in the gas (whose temperatureincreases consequently) and could be recovered at the endof the link. Alternatively, losses could be used for increas-ing the gas temperature before expanding it to keep thetemperature within required limits. Comparing such dualconcepts with conventional, decoupled transmission linesshows advantages and disadvantages.

From an energetic point of view, combined transmis-sion is more efficient if the heat losses can be used at theend of the link. From a legal point of view, the devicecould be interesting since rights of way and other issuescould be managed for electrical and chemical transmissionsimultaneously. Like normal pipelines, the energy inter-connector can also be used for gas storage (line pack). Anissue to consider is the dependability of the interactingpower flows (electricity and gas), which could reduce sup-ply redundancy. Considering contingencies on the onehand, common mode failures could be a serious issue. Onthe other hand, investigations have shown that operationalboundaries arise from the coupling of the flows. Simplyspeaking, a certain gas flow is necessary to provide suffi-cient cooling for the electrical conductor. Studies haveshown that these operational restrictions can be relievedwhen combining energy interconnectors with energy hubs.However, under certain circumstances, the energy inter-connector promises better performance than traditional,separated transmission technologies. The integration ofgaseous and electrical energy transmission is only one of

several possible approaches. Concepts involving liquidchemical carriers or further forms of energy may beadvantageous as well.

New Models and Analysis ToolsEconomic and physical performances of different energycarriers are well understood, but global features of inte-grated systems have not yet been investigated extensively.Since there are only a few tools available for the analysisof such systems, the development of a modeling andanalysis framework for multicarrier energy systems has

27

figure 3. Possible layout of an energy interconnector.

Electricity

Gaseous Hydrogen

Heat

Losses

been identified as an essential need. The aim was todevelop the same tools as are available for electricity sys-tems—e.g., power flow, economic dispatch, reliability,and stability. In addition, models for storage and intercon-nector technology were developed that are capable ofintegrating into the system analysis framework.

Power FlowFor general investigations on the system level, steady-state flow models are appropriate and commonly used.The flows through power converter devices is simply ana-lyzed, defining their energy efficiency as the ratio ofsteady-state output and input. With multiple inputs andoutputs, a conversion matrix can be defined that links thevectors of the corresponding power flows. Figure 4 out-l ines this modeling concept. The coupling matrixdescribes the transformation of power from the input tothe output of the hub; it can be derived from the hub’sconverter structure and the converters’ efficiency charac-teristics. Describing the behavior of storage devicesrequires considering time and energy as additional vari-ables. Various flow models are available for hydraulic andelectric networks, from general network flow to moredetailed steady-state power flow models. The appropriatedegree of approximation depends on the kind of investiga-tion. Combined transmission links (interconnectors) canbe modeled similar to energy hubs via coupling matrices.

ReliabilityReliability and availability of energy supply is an importantdesign criterion; therefore, models have also been developedfor this kind of investigation. Failure and repair rates can bedefined for all components in the system. Considering anenergy hub, failure and repair rates of the coupling elements

can be stated in matrices similar to theconversion matrix. The influence of theenergy hub, i.e., an increase or decreaseof availability between input and outputof the hub, can be analyzed with thisapproach. Furthermore, the model can beused in the optimization process.

System OptimizationVarious optimization problems can beidentified when considering integratedmulticarrier systems. The basic ques-tion of combined optimal power flowis how much of which energy carriersthe hubs should consume and how theyshould be converted in order to meetthe loads at their outputs. This is anoperational problem. In the planningphase, the optimal structure of the hubmay be of interest, which can be foundby determining the optimal coupling

matrix that describes the conversions within the hub. Con-verters can then be selected to establish this optimal cou-pling, and missing technologies can be identified. Theseand other optimization problems have been formulatedand analyzed using various criteria such as energy costs,system emissions, and transmission security measures.Multiobjective optimization can be performed by combin-ing different criteria in composite objective functions.

28 IEEE power & energy magazine january/february 2007

figure 5. Result of investment analysis. The present valueof the device (per MW electrical output) increases with itstotal efficiency, since more energy cost are saved in eachperiod if the efficiency is higher. Today’s investment cost forCHP units of comparable size are in the range of €500,000per MW electrical output (rated). The conclusion that canbe drawn from this plot is that an investment is reasonable ifa total efficiency of more than 75% can be achieved.

figure 4. Modeling the transformation of power through an energy hub.

Energy HubOutput Energy

Carriersα, β, ...ζ

Input Power VectorOutput Power Vector Coupling Matrix

α

β

ζζ

α

β

ζζ

Input EnergyCarriersα, β, ...ζ

Poutα

Poutβ

Pout

Cαα

Cαβ

Cαζ

Cβα

Cββ

Cβζ

Cζα

Cζβ

Cζζζ

Pinα

Pinβ

Pinζ

=

Pre

sent

Val

ue in

k /

MW

e

Total Efficiency in %

70 80 90

482

675

824

RealisticToday

january/february 2007 IEEE power & energy magazine

Evaluation of InvestmentWhen talking about completely new systems on the greenfield, the question of cost plays one of the most importantroles. Energy prices and savings in energy cost can be esti-mated, although assumptions are often critical. The evalua-tion of investment costs is more difficult. How much willnew technologies such as fuel cells costs in 30–50 years?To avoid speculations based on doubtful assumptions, thequestion is put differently. The justifiable investment costsare determined by comparing the performances of the con-ventional and the proposed/assumed system. For example,energy cost and CO2 taxes can be compared for a conven-tional system and an optimized greenfield structure. Fromthe annual savings due to higher energy efficiency and lessemissions of new technologies, a present value can bedetermined that represents the break-even investment costof the new technology. With this method, results stilldepend on critical assumptions as inflation, compounding,and risk. However, using this tool for sensitivity analysisyields deeper insight into economics; it enables identifica-tion of the significant parameters.

Figure 5 shows an example where the sensitivitybetween total energy efficiency ofa cogeneration-equipped energyhub and its justifiable investmentcost was determined. In this partic-ular case, results show that evenstate-of-the-art technology couldkeep up with the requirements, i.e.,install ing such cogenerationdevices would be reasonable froman economic point of view (undercertain assumptions).

A First ApplicationThe energy hub idea was picked upby a municipal utility in Switzer-land, the Regionalwerke AG Baden,which plans to build an energy hubcontaining wood chip gasificationand methanation and a cogeneration plant. The idea is togenerate synthetic natural gas (SNG) and heat from woodchips, a resource which is available in the company’s sup-ply region. The produced SNG can then either be directlyinjected in the utility’s natural gas system, or convertedinto electricity via a cogeneration unit and fed into the

electric distribution network. Waste heat, which accrues inboth cases, can be absorbed by the local district heatingnetwork. The whole system can be seen as an energy hubprocessing different energy carriers—wood chips, electric-ity, heat, and SNG. In addition to these energy carriers, thegasification process requires nitrogen and steam, whichhave to be provided at the hub input. Figure 6 gives anoverview of the hub layout. The new thing here is not thetechnology used (converters), but the integrated planningand operation, which is believed to enable better overallsystem performance.

The developed multicarrier analysis tools can be appliedto this energy hub to answer some fundamental questions.

✔ Design/Dimensioning: How should the converters berated, i.e., how much electricity, SNG, and heat shouldthe hub be able to produce?

✔ Operation: How should the energy hub be operated,how much electricity/SNG/heat should be generateddepending on the actual load situation?

✔ Storage: Which and how much of which energy carriershould the energy hub be able to store—wood chips,SNG, heat, electricity?

✔ System Impact: How does the energy hub influence theoverall system performance in terms of reliability/availability, energy efficiency, and power quality?

The project is still in the planning phase. A first version ofthe hub, which only contains wood gasification and cogener-ation units, should be realized by 2009. The full version,

29

figure 6. Sketch of the energy hub to be realized by Regionalwerke AG Baden,Switzerland.

Synergy effects among various energy carriers can be achieved by taking advantage of their specific virtues.

Electricity

Wood Chips

Steam

Nitrogen

Electricity

Heat

SNG

Methanation

Cogeneration

GasificationDrying

Energy Hub

which then includes the methanation part (thus enablinginfeed of SNG into the natural gas network) should start run-ning in 2011.

ConclusionsThe research project “Vision of Future Energy Networks”distinguishes itself from others by aiming at a greenfieldapproach, integrating multiple energy carriers, and consider-ing a timeframe of 30–50 years from now. The definition ofenergy hubs and the conception of combined interconnectordevices represent key approaches towards a multicarriergreenfield layout. Models and tools for technical (e.g., powerflow, reliability), economical (e.g., energy and investmentcost), and environmental (e.g., CO2 emissions) investigationsin multicarrier energy systems have been developed and usedin various case studies. The main conclusions that can bedrawn so far are as follows.

✔ The energy hub concept enables new design approach-es for multiple energy carrier systems.

✔ The flexible combination of different energy carriersusing conversion and storage technology offers a pow-erful approach for various system improvements. Ener-gy cost and system emissions can be reduced, securityand availability of supply can be increased, congestioncan be released, and overall energy efficiency can beimproved.

✔ The developed modeling and analysis framework pro-vides suitable tools for the planning and operation ofmultiple energy carrier systems.

Future work includes the development of dynamic modelingand analysis tools (e.g., for evaluating stability), and the con-trol of a system of interconnected energy hubs (centralizedversus decentralized, agent-based). The concepts will be fur-ther refined and elaborated in more detail using realisticexamples and case studies.

AcknowledgementsFinancial support from ABB, Areva, Siemens, and the SwissFederal Office of Energy is gratefully acknowledged.

For Further Reading“Vision of future energy networks,” project home page, 2006,[Online]. Available: http://www.eeh.ee.ethz.ch/psl/research/vofen.html.

P. Favre-Perrod, M. Geidl, G. Koeppel, and B. Klockl,“A vision of future energy networks,” in Proc. IEEE Inau-

gural Conf. Exposition Africa, Durban, South Africa, 2005,pp. 13–17.

M. Geidl, P. Favre-Perrod, B. Klockl, and G. Koeppel, “Agreenfield approach for future power systems,” in Proc. CigreSession 41, Paris, France, 2006, CD-ROM, C6-307.

BiographiesMartin Geidl received a Dipl.-Ing. degree in electrical engi-neering from the Graz University of Technology, Austria, in2003. After that he joined the Power Systems Laboratory ofETH Zurich, Switzerland, where he is working towards aPh.D. He is a Student Member of the IEEE.

Gaudenz Koeppel received his M.S. in electrical engineer-ing from ETH Zurich, Switzerland, in 2002. He is currentlypursuing his Ph.D. degree at the Power Systems Group atETH Zurich. He is a Student Member of the IEEE.

Patrick Favre-Perrod graduated from ETH Zurich,Switzerland, in 2003. He is currently working on a thesisabout energy transmission in future power systems at the highvoltage laboratory of ETH Zurich. He is a Student Memberof the IEEE.

Bernd Klöckl received his MSc degree from Graz Univer-sity of Technology, Austria, in the areas of electric machinesand high voltage technology in 2001. Since 2002, he hasbeen with the High Voltage Laboratory of ETH Zurich,Switzerland. He is a Student Member of the IEEE.

Göran Andersson obtained his M.Sc. and Ph.D.degrees from the University of Lund in 1975 and 1980,respectively. In 1980, he joined ASEA, now ABB, HVDCdivision in Ludvika, Sweden, and in 1986, he wasappointed full professor in electric power systems at theRoyal Institute of Technology (KTH), Stockholm, Swe-den. Since 2000, he has been a full professor in electricpower systems at ETH Zurich, Switzerland, where heheads the Power Systems Laboratory. He is a Fellow ofthe IEEE and chairs the IEEE PES Power SystemDynamic Performance Committee.

Klaus Fröhlich received the M.Eng. and Ph.D. degreesin technical science from the Vienna University of Technol-ogy, Austria. After 11 years in switchgear and high-voltagetechnology with BBC (later ABB) in Switzerland, hebecame a full professor at the Vienna University of Tech-nology in 1990. Since 1997, he has been a full professor ofHigh Voltage Technology at the Swiss Federal Institute ofTechnology Zurich, Switzerland. He is a Fellow of theIEEE and chairs the Cigre Technical Committee.

30 IEEE power & energy magazine january/february 2007

An energy hub is considered a unit where multiple energy carriers can be converted, conditioned, and stored.

p&e