Business ecosystem modeling- the hybrid of system modeling and ecological modeling: an application of the smart grid

System Modeling Process

Systems are complex, and system modeling approaches are different based on the aspects/layers of the systems required to be modeled. In system engineering, there are several well-known system modeling processes, e.g. the ISO 15288 (Industrial automation systems and integration- Integration of life-cycle data for process plants including oil and gas production facilities) (ISO 2013) and the system engineering formula by NASA (2007). In the ISO 15288, there are 25 processes defined to realize a system, starting from a statement of purpose (objective) and a set of top-level (stakeholder) requirements ending in operating and maintaining the system (Ruijven 2012). Meanwhile, NASA creates the single systems engineering formula (NASA n.d.) as:

SE (System Engineering) = Vee (the V shape diagram, shown in Fig. 1) + 11 SE Functions (shown in Fig. 2) + Tools.

Open image in new window

Open image in new window

Architecture development is an important part of system engineering because a system architecture is a conceptual model that defines the structure, behavior, and multiple views of a system (Jaakkola and Thalheim 2011). Among a number of the architecture development approaches, the TOGAF (The Open Group Architecture Framework) (The Open Group n.d.-a) and DoDAF (the Department of Defence Architecture Framework) (Chief Information Officer n.d.) are popularly used as the standards for the enterprise architecture (Tao et al. 2017).

Domain Ontology

An ontology is an inventory of the kinds of entities that exist in a domain, their salient properties, and the salient relationships that can hold between them (Benjamin et al. 1995). In a domain ontology, various kinds of objects (e.g., tools and employees), properties (e.g., being made of metal or professional expertice), and relations between tools and employees (e.g., produced by or working with) are defined (Benjamin et al. 2006).

Ontologies are important in intelligent system development for modeling interoperability, composition, and information exchange at the semantic level. Ontologies addresses four challenges at the semantic level: semantic inaccessibility, logical disconnectedness, consistency maintenance, and modeling at multiple levels of abstraction (Benjamin and Graul 2006). However, not all ontologies represent domains in the same way and at the same level of detail. In addition, ontologies use different vocabularies to describe the same concepts. Therefore, there is the need for a unified ontology as a standard that can be used for a specific domain, e.g. smart grid (Cuenca et al. 2017).

For a domain ontology development, Step 1 is very important that determines the accuracy and detail of domain analysis and description. For instance, (Cuenca et al. 2017) investigates the ontologies of the energy data domain, and finds that none of these ontologies represents all types of energy performance and contextual data that should be taken into account for energy management and the data are also represented at different levels of detail (e.g. Table 2). To reuse existing ontologies, it is necessary to have a clear picture of the ontology network for the interconnected domains that related to the domain determined at Step 1 (e.g. Fig. 3). Then which parts of the existing ontologies and what concepts should be selected and reused can be decided (e.g. Table 3).

Table 2

Energy Domains Representation Level of Detail (Cuenca et al. 2017)

Energy domain	Ontology
	Think Home	SAREF4EE	BOnSAI	ProSGV3
Infrastructure technical data	H	L	M	M
Energy consumption systems data	H	M	L	H
Energy performance data	H	H	H	H
Sensors/actuators data	H	M	M	M
Energy stakeholders’ data	M	–	L	L
Weather/climate data	H	L	L	M
Geographical data	–	–	–	L
Environmental data	M	–	M	–
Distributed energy sources data	M	L	L	M
Energy DR operations	–	M	–	L

(H=High/M = Medium/L = Low)

Open image in new window

Table 3

Example of Reused Ontologies and Concepts (Cuenca et al. 2017)

Energy domain	Ontology	Reused concepts
Energy consumption systems data	ThinkHome ontology	HVAC systems, communication appliances, entertainment appliances, oce automation devices, lighting systems, white goods, acoustic systems, domotic network components, energy facilities, device state, device commands, equipment manufacturer, external and internal equipment
	SAREF4EE ontology	Appliances working modes and power proles, device manufacturer and device model.
	EnergyUse ontology	Wearable devices.
	ProSGV3 ontology	Body care devices, pressing devices, water heating devices, charging devices, lighting systems, entertainment devices, cleaning devices and electrical appliance category.

Ontologies are well discussed in the engineering domain, especially in IT & software engineering. Although there are ontologies in the business domain, the business ontology development is not popular. One of the best known business ontologies is the business model canvas, an ontology for business models (Osterwalder 2004). Another business ontology that is popularly applied in enterprise architecture modeling is the SOA (Service-Oriented Architecture) ontology (The Open Group 2014) that is part of TOGAF (The Open Group n.d.-a). The class hierarchy is shown in Fig. 4, and shows that the SOA ontology can contribute to the information flow, improve alignment between the business and information technology communities, and facilitate SOA adoption (The Open Group 2014).

Open image in new window

System Modeling Languages and Tools

The UML class diagramming has been popularly used for the system modeling. It has been used in the smart grid system modeling for modeling the electricity market (e.g. the harmonized role model (E. a. e. ENTSO-E 2018)), IT infrastructure and multi-layer architecture (in the SGAM (Smart Grids Architecture Model) framework (CEN-CENELEC-ETSI Smart Grid Coordination Group 2012)). The use case diagram is also used in the SGAM framework to present the design of the smart grid in an architectural viewpoint that allows both- specific but also neutral regarding solution and technology. SGAM framework also uses the sequence diagrams to describe the information objects which are exchanged between actors.

The SGAM framework provides a guide for the smart grid architecture, and it does not directly apply the SysML (Systems Modeling Language) for the architecture design. SysML is well known as a general-purpose architecture modeling language for system engineering applications (shown in Fig. 5) that is adopted by the Object Management Group (OMG) in 2006 (SysML.org 2019). SysML is defined as an extension of a subset of the UML using UML’s profile mechanism (SysML Forum n.d.) (shown in Fig. 6). There are a number of SysML Modeling Tools for the specification, analysis, design, verification and validation of a broad range of systems and systems-of-systems (SysMLtools.com n.d.).

Open image in new window

Open image in new window

Biological Ecosystem

The term ecosystem was firstly used by A. G. Tansley in 1935, and he defined an ecosystem as ‘a particular category of physical systems, consisting of organisms and inorganic components in a relatively stable equilibrium, open and of various sizes and kinds’ (Tansley 1935). Later, the definitions of ecosystem emphasize more on ‘interaction’. For instance, the ecosystem is defined as the joint functioning and interaction (large quantities of matter, energy and information) of these two compartments (populations and environment) in a functional unit of variable size (Garcia, and Food, and A. O. o. t. U. Nations 2003).

According to (Ricard 2014), ‘the “species” is the ecosystem basis and refers to all organisms of the same kind. The members of a species living in a given area at the same time constitute a population. All the populations living and interacting within a particular geographic area make up a biological (or biotic) community. The living organisms in a community together with their non-living or abiotic environment make up an ecosystem’.

Based on the scale of ecological dynamics, ecosystems hierarchy includes ecozone, biome, ecosystem (ecological community, population, habitat/biotope), ecotone and niche (Ricard 2014).

Ecosystems are controlled both by external and internal factors, e.g. climate, soil, topography. External factors control the overall structure of an ecosystem and the way things work within it. Internal factors control ecosystem processes and are controlled by the feedback loops. Internal factors include decomposition, competition or shading, disturbance, succession and the types of species present.

Ecosystem science is the study of inter-relationships among the living organisms, physical features, biochemical processes, natural phenomena, and human activities in ecological communities (NOAA 2018). Ecosystem science investigates the areas (not limited to) of ecosystem processes, population ecology and population dynamics, disturbance and resilience, nutrient cycle, decomposition and mineralization, ecological amplitude, ecology, environmental influences, biological interactions, biodiversity, and environmental degradation. According to (Gorham and Kelly 2018), types of ecosystems, environmental factors, levels of complexity, major plants and animals, and interspecific associations have been popularly discussed in the ecosystem science.

The theories of the biological ecosystem demonstrate that the terminologies, principles and methodologies in the biological ecosystem science can be modified and applied in other disciplines, e.g. system engineering and business ecosystem research.

Ecosystem Mapping

Ecosystem mapping is the spatial delineation of ecosystems following an agreed ecosystem typology (ecosystem types), which strongly depends on mapping purpose and scale (European Union 2013). The ecosystem mapping may also include mapping of status (e.g. functioning and health) for monitoring and assessment of the ecosystem’s quality.

According to a coherent analytical framework proposed by EU ( 2013), the ecosystem classification and mapping apply two basic principles: typological (divides nature into ecosystem types – classes that can occur at more geographical locations, e.g., temperate broadleaf and mixed forests) and regional (describes ecosystems from a regional perspective e.g., Dinaric mixed forests) or their combination.

Ecosystem services are an important part of the ecosystem mapping that describe the ways that humans are linked to and depend on nature. There are different definitions of ecosystem services ((La Notte et al. 2017) summarizes the definitions provided in recent ecosystem services studies). For instance, MA (The Millennium Ecosystem Assessment) defines ecosystem services as “the benefits people obtain from ecosystems” (Millenium Ecosystem Assesment 2005), and ecosystem services are defined as the contributions that ecosystems make to human well-being by CICES (the Common International Classification for Ecosystem Services) (European Union 2013). The conceptions of ecosystem services have been discussed in different domains, e.g. natural science, economics and policy (Danley and Widmark 2016).

There are three international classification systems available to classify ecosystem services: MA (Millennium Ecosystem Assessment 2005), TEEB (The Economics of Ecosystems and Biodiversity n.d.) and CICES (n.d.). Ecosystem services categories in CICES are provisioning services, regulating and maintenance services, and cultural services (Haines-Young and Potschin 2018). CICES does not include the ‘supporting services’ in MA, but merges the ‘habitat services’ from TEEB with regulating services, in a category called ‘regulating and maintenance services’ (La Notte et al. 2017).

Compared to others, CICES promotes a clear distinction between ecosystem services and ecosystem benefits, and can help negotiate the different kinds of perspective that have developed around the ecosystem service concept and assist in the exchange of information about them (Haines-Young and Potschin 2012). CICES describes ‘final ecosystem services’ using a five-level hierarchical structure, and the way the system works can be illustrated in Fig. 7 (CICES n.d.). Meanwhile, the ecosystem service cascade also presents the relations between the environment and the social & economic system, and services can have social as well as economic value (Haines-Young and Potschin 2013).

Open image in new window

Compared to CICES that places greater emphasis on the ecological system, FEGS-CS (Final Ecosystem Goods and Services Classification System) (United States Environmental Protection Agency (EPA) n.d.) and NESCS (the National Ecosystem Services Classification System) (Landers 2015) mainly focus on benefits and beneficiaries (described in (Landers and Nahlik 2013)). Based on the comparison between CICES and FEGS-CS (La Notte et al. 2017; Bordt 2016), this paper recommends including benefits and beneficiaries in the business ecosystem modeling because benefits and beneficiaries are an important aspect in the business ecosystem. However, there is a lack of discussion regarding whether CICES and FEGS-CS can be compatible. The Swiss Federal Office for the Environment (FOEN) proposes an ‘Inventory of Final Ecosystem Goods and Services’ (FEGS) and include four Benefit categories (health, security production factors and natural diversity) (Staub et al. 2011). The FEGS aims to be integrated into the MA and CICES Classifications.

Ecosystem Modeling

Ecosystem modeling is also called ecological modeling. An ecosystem model is an abstract representation of an ecological system which is studied to better understand the real system (Hall and Day 1990). Ecological modeling can be simply categorized as conceptual models and quantitative models. Generally, conceptual models are written as diagrams with boxes (represent state variables or condition of the ecosystem components) and arrows (illustrate relationships among state variables), and a quantitative model is a set of mathematical expressions of coefficients and data attached to the boxes and arrows in the conceptual models (Jackson et al. 2000). Quantitative models are more precise and specific about a system and can make predictions when they capture the key elements for system representation (Bondavalli et al. 2009). There are three main types of ecosystem modeling and the choice of model type and detail depends on the system studied, the questions asked, and the data available (Jackson et al. 2000).

Besides agent-based modeling, there is an increase in applying multi-agent simulation in ecology due to the growth in CPU power. In the field of ecosystem management, the interactions between ecological dynamics and social dynamics are examined, and modelers describe systems as a set of modules or compartments interlinked by flows (of matter, energy, or information) and controls (Bousquet and Le Page 2004). Multi-agent systems are used in ecology to investigate multiple agent interactions (Huhns and Stephens 1999). Multi-agent system originally came from the field of artificial intelligence, and is a complex system that is composed by more than one distributed agents, and these agents communicate to deal with problems which usually can’t be solved by a single agent (Dam and Lukszo 2006; Merdan et al. 2011).

Smart Grid Architecture

The concept of the smart grid is built upon the intersection of technology, people and infrastructure for intelligent generation, transmission, distribution and consumption of electricity (Geelen et al. 2013). A smart grid can be defined in terms of the socio-technical network with enhanced two-way communication and active management of information and energy flows that allow controlling the practices of distributed generation, storage, consumption and flexible demand (Wolsink 2012).

There are many different definitions of a smart grid that emphasizes different aspects (Ma et al. 2015). To have a holistic view of the smart grid with multiple stakeholders, multiple applications, multiple networks, it is necessary to investigate the smart grid architecture. There are several smart grid models have been proposed and discussed, e.g. the smart grid conceptual model introduced by The National Institute of Standards and Technology (NIST) (2018) and SGAM Framework developed by CEN (the European Committee for Standardization), CENELEC (the European Committee for Electrotechnical Standardization), and ETSI (the European Telecommunications Standards Institute) (CEN-CENELEC-ETSI Smart Grid Coordination Group 2012).

Compared to other smart grid models or framework, the SGAM Framework can give an overall picture of the smart grid with the description of the major elements, integrate other models and architectural frameworks, e.g. GWAC (GridWise Transactive Energy Framework) (The GridWise Architecture Council 2015) and TOGAF (The Open Group n.d.-a), etc. The SGAM framework spans three dimensions: domain, interoperability (layer) and zone (shown in Fig. 8, and explained in Table 8), and it visualizes the relations among these three dimensions with multi-layers.

Open image in new window

Table 8

Table Styles Terms and Definitions in the SGAM Framework (CEN-CENELEC-ETSI Smart Grid Coordination Group 2012)

Terms	Definitions
SGAM Interoperability Layer	To allow a clear presentation and simple handling of the architecture model, the interoperability categories described in the GridWise Architecture model are aggregated in SGAM into five abstract interoperability layers.
SGAM Domain	One dimension of the Smart Grid Plane covers the complete electrical energy conversion chain, partitioned into 5 domains.
SGAM Zone	One dimension of the Smart Grid Plane represents the hierarchical levels of power system management, partitioned into 6 zones.

Smart Grid Actors and Roles

The harmonized role model (E. a. e. ENTSO-E 2018) proposed by ebIX, EFET and ENTSO-E define each market role related to all the different electricity markets in Europe. An example of the roles and description in the harmonized role model is shown in Table 10. UML class diagramming is used in the harmonized role model, that the “actor” symbol is used to represent a role and the “class” symbol is used to define a domain (shown in Fig. 9).

Table 10

An Example of the Roles and Descriptions in the Harmonized Role Model (E. a. e. ENTSO-E 2018)

Role	Description
System operator	Has overall responsibility for creating balance in the market and for handling transmission grid operation and ensuring stable electricity supply.
Grid Operator	A party that operates one or more grids
Meter Operator	A party responsible for installing, maintaining, testing, certifying and decommissioning physical meters.

Open image in new window

However, the definition of a business actor corresponds 1:1 with a role (E. a. e. ENTSO-E 2018; IEC 2012) does not comply in all smart grid system. Unbundling is defined by EU (European Commission n.d.) as ‘the separation of energy supply and generation from the operation of transmission networks’. Comparatively, in the bundling electricity market structure, the energy supply and generation might not be separated from the operation of transmission network, and one actor can hold one or more than one roles. For instance, the actor-utility in the Chinese electricity market corresponds to the roles of transmission operator, distribution operator, and electricity supplier.

These definitions can better represent the relations between actors and roles in a smart grid.

Value Flows in the Smart Grid

Besides the emphasis on actors and roles in both the business layer in the SGAM framework and in the conceptualization of the business ecosystem, interactions between actors are also important and well discussed in both the SGAM framework specification and business ecosystem litterature.

For instance, in the SGAM conceptual model, actors are connected by associations, the exchanged information is defined as business services, and actors interact through these business services. However, the description of the associations or business services is not detail discussed in the SGAM framework. Meanwhile, although processes/ interactions between actors are described in the harmonized role model (E. a. e. ENTSO-E 2018) and IEC standards (e.g. 61,850) (Energinet 2019), the described interactions in (E. a. e. ENTSO-E 2018; Energinet 2019) are mainly technical related information, and there is no methodology for systemically defining and describing the interactions between actors in the smart grid.

In the theory of the business ecosystem, especially the service ecosystem, value co-creation is well discussed for the understanding of interactions between actors, e.g. (Palumbo et al. 2017). Value co-creation is driven by the collaborative efforts of and interactions between actors in the ecosystem (Pop et al. 2018). Value can be tangible and intangible (Ciasullo et al. 2017), and at different levels: micro (e.g., households, organizations, etc.), meso (e.g., industries, communities, etc.), and macro (e.g., nations, global markets, etc.) (Frow et al. 2016). However, the types of value are not defined in the ecosystem theory, but value constitutesis an important part of the ecosystem modeling to define and clarify the interactions between actors.