=Paper=
{{Paper
|id=Vol-1501/Diversity2015-paper_4
|storemode=property
|title=An Ontology For Specifying Spatiotemporal Scopes in Life Cycle Assessment
|pdfUrl=https://ceur-ws.org/Vol-1501/Diversity2015-paper_4.pdf
|volume=Vol-1501
|dblpUrl=https://dblp.org/rec/conf/semweb/YanHKJBKJHSI15
}}
==An Ontology For Specifying Spatiotemporal Scopes in Life Cycle Assessment==
https://ceur-ws.org/Vol-1501/Diversity2015-paper_4.pdf
An Ontology For Specifying Spatiotemporal Scopes in
Life Cycle Assessment
Bo Yan1 , Yingjie Hu1 , Brandon Kuczenski1 , Krzysztof Janowicz1 , Andrea Ballatore1 ,
Adila A. Krisnadhi2,4 , Yiting Ju1 , Pascal Hitzler2 , Sangwon Suh1 , Wesley Ingwersen3
1
University of California, Santa Barbara, USA
2
Wright State University, USA
3
US Environmental Protection Agency, USA
4
University of Indonesia, Indonesia
Abstract. Life Cycle Assessment (LCA) evaluates the environmental impact of
a product through its entire life cycle, from material extraction to final disposal
or recycling. The environmental impacts of an activity depend on both the ac-
tivity’s direct emissions to the environment as well as indirect emissions caused
by activities elsewhere in the supply chain. Both the impacts of direct emissions
and the provisioning of supply chain inputs to an activity depend on the activity’s
spatiotemporal scope. When accounting for spatiotemporal dynamics, LCA often
faces significant data interoperability challenges. Ontologies and Semantic tech-
nologies can foster interoperability between diverse data sets from a variety of do-
mains. Thus, this paper presents an ontology for modeling spatiotemporal scopes,
i.e., the contexts in which impact estimates are valid. We discuss selected axioms
and illustrate the use of the ontology by providing an example from LCA practice.
The ontology enables practitioners to address key competency questions regard-
ing the effect of spatiotemporal scopes on environmental impact estimation.
Keywords: Life cycle assessment; ontology; spatiotemporal scopes.
1 Introduction
Life Cycle Assessment (LCA) is a method for analyzing the environmental impact of a
product or a service through all stages of its life cycle [5, 13]. LCA is designed to take
into consideration the entire life cycle and product chain, having a holistic viewpoint in
dealing with environmental issues. The life cycle of a product or a service normally in-
cludes raw material acquisition, manufacturing process, trading process, product usage,
recycling process, and waste management. Due to the diverse types of information re-
quired to conduct an LCA, knowledge from different domains needs to be gathered and
interpreted together. This process is challenging because there is no universal ontology
or vocabulary among (or even within) domains of study, creating substantial barriers
for information sharing and integration among different data providers.
An important stage of LCA is the creation of a Life Cycle Inventory (LCI), which
consists of representing economic activities as a collection of unit processes linked to-
gether through interdependency relations [3]. A global understanding of these economic
activities and the data that comes along require a clear capture of the interdependencies
and relationships between the used nomenclatures. Semantic technologies and ontolo-
gies are promising methods to support interoperability in LCA. They foster semantic
interoperability without the need to enforce a single domain schema.
2
Researchers have long been considering the significance of a spatial perspective in
LCA [11, 2, 6]. Because environmental impacts in LCA are driven by the emission of
substances into the environment, site-specific assessments are often necessary to deal
with spatial variation, which refers to differences in geology, topography, land cover,
and so forth [12]. The geographic location of a process can also be important in de-
termining the impacts of activities that occur elsewhere in the supply chain, such as
freight transport. Likewise, a number of studies have argued that a dynamic approach
is necessary to account for temporal variations in activities or impacts [9, 14]. These
investigations show that the spatiotemporal scope of assessed activities is important
to LCA. While recent studies [3, 18] have used semantic technologies for LCA, most
of them are too general and have not taken scoping into account. Here, we introduce
a compact ontology that formalizes the spatiotemporal scope of activities in LCA and
integrates well with previously published LCA-related ontology design patterns [8, 15].
While our work is concerned with LCA as a diverse field that will benefit from the
ontological modeling of its data and workflows, spatiotemporal scopes are relevant for
a multitude of other domains. Thus, the core part of our ontology can be regarded as an
ontology design pattern (ODP).
2 Competency Questions
Designing an ontology requires generic use cases to capture the recurring problems in
one or multiple domains. Competency questions have been recognized as an effective
approach to identify such use cases. Competency questions are frequent queries that
subject matter experts would like to submit to a knowledge base to find answers. The
following listing shows examples of competency questions that have been identified by
international LCA experts during the GeoVoCamp Santa Barbara 2015:
– Question 1: "What is the emission of activity a at place p at time t?"
– Question 2: "What are the supply chain requirements of activity a when it happens
at place p1 and time t1 ?"
– Question 3: "What is the difference of activity a at places p1 and p2 at the same
time t?"
– Question 4: "What is the difference of activity a at times t1 and t2 at the same
place p?"
As can be seen, answering the four competency questions requires four main con-
cepts: Activity, Flow, Place, and Time. An activity in LCA may have Requirement
(Question 2) and Outcome (e.g., emission) (Question 1), which collectively make up
flows. To effectively link these concepts, proper relations have also been specified.
3 Spatiotemporal Scoping Ontology
The ontology for spatiotemporal scopes5 is developed based on the existing Activity
ODP, [1] which also includes concepts such as Activity, Requirement, and Outcome.
Since the Activity ODP focuses on human activities, we modify and extend it to fit
5
http://descartes-core.org/ontologies/lca/1.0/stscope.owl
3
Fig. 1. An overview of the spatiotemporal scope ontology.
the activities in the domain of LCA and add the scoping on top of the resulting model.
Figure 1 provides an overview of the spatiotemporal scoping ontology. In the following,
we describe the classes and relations by showing selected description logics axioms.
SpatiotemporalScope: This class represents the spatial and temporal context un-
der which an LCA activity occurs. We define this class instead of directly using Place
and Time, because many requirements and outcomes are associated with the spatiotem-
poral contexts rather than places and time (intervals) alone. SpatiotemporalScope is
associated with the classes Place and Time through the relations occursAtPlace and oc-
cursAtTime. Time should be an interval because an LCA activity generally represents
the performance of a typical facility or a set of facilities over a time period, rather than
at a specific moment. A place is some, typically named, extent in geographic space,
e.g., a country or region. We do not specify both classes here and refer to OWL-Time
and GeoSPARQL for details (which implies that geometry types such as multi-part pol-
gyons can be used as spatial footprints of places). We assert that each spatiotemporal
scope has at least one place and time (see Eq. 1).
SpatiotemporalScope v ∃occursAtP lace.P lace u ∃occursAtT ime.T ime
u ∃hasSpatiotemporalScope− .Activity (1)
Note that we also provide role chains in the ontology (see Eq. 2 as example) but do not
discuss them here for lack of space.
hasSpatiotemporalScope ◦ occursAtP lace v scopedByP lace (2)
hasSpatiotemporalScope ◦ occursAtT ime v scopedByT ime (3)
Activity: The Activity class represents activities in the LCA sense and thereby differs
from other conceptualizations of activities. An activity roughly corresponds to a unit
process as defined in the ISO 14044 standard [7], but may also indicate a reservoir,
stock, or natural process such as dissolution into fresh water. An activity always occurs
at a certain place (e.g. in a particular factory or river) in a particular time span. We
link Activity to SpatiotemporalScope through the relation hasSpatiotemporalScope. An
activity in LCA also has Requirement and Outcome.
Activity v ∃hasRequirement.Requirement u ∃produces.Outcome (4)
u ∃hasSpatiotemporalScope.SpatiotemporalScope . . .
4
Flow: A Flow is a highly generic concept in LCA mainly defined as a counterpart to
an activity. A flow may represent the transfer of matter, such as an emission of combus-
tion gases, or an exchange of services, such as transporting a good. A flow is exchanged
between an activity and another activity. Although flows are the products of processes,
many flows can exist independently of any process, can be accumulated in reservoirs,
and can have properties, such as economic value. A flow can play a role in both the
Requirement of one activity and Outcome of another, and thus, we define the relation of
hasRole. If the two activities in the exchange are industrial unit processes, the flow is
referred to as an intermediate flow. The two activities can be described as “partners” to
the exchange. A flow exchanged with the natural environment is called an elementary
flow. We formalize the class of Flow as below:
F low v ∃hasRole.(Requirement t Outcome) (5)
An ontology for flows has recently been developed and formally specifies the distinc-
tions made above [8].
Requirement and Outcome: Any activity in LCA has required inputs and resulting
outputs. These inputs and outputs are formalized as Requirement and Outcome in our
ontology. The provision of Requirements and the disposition of Outcomes depend on
the specific place and time that an activity takes place. Thus, we define Spatiotemporal-
Requirement as a subclass of Requirement, and associate it to the SpatiotemporalScope
using the relation isAssociatedWith. Similarly, we define SpatiotemporalOutcome as a
subclass of Outcome, and use isInfluencedBy to link it with SpatiotemporalScope.
SpatiotemporalRequirement v Requirement (6)
SpatiotemporalOutcome v Outcome (7)
SpatiotemporalRequirement v ∃isAssociatedW ith.SpatiotemporalScope (8)
SpatiotemporalOutcome v ∃isInf luencedBy.SpatiotemporalScope (9)
Domain & Range Restrictions and Class Disjointness: In addition to the above
axioms, the pattern also defines a set of guarded domain and range restrictions. Specif-
ically, for each object property P pointing from the class A to the class B in Figure
1, we define ∃P.B v A as the guarded domain restriction and A v ∀P.B as the
guarded range restriction, which also acts as a local closure of P . For example, for
occursAtP lace property, we have:
∃occursAtP lace.P lace v SpatioT emporalScope (10)
SpatioT emporalScope v ∀occursAtP lace.P lace (11)
Specific for hasRole property, we have:
∃hasRole.Requirement t ∃hasRole.Outcome v F low (12)
F low v ∀hasRole.(Requirement t Outcome) (13)
Finally, we assert class disjointness for every pair of classes in Figure 1, except when
the pair of classes are connected via rdfs:subClassOf.
5
4 Use Case
A use case was created by selecting a unit process from the US Life Cycle Inventory
database, in this case “Conditioned log, at plywood plant, US SE,” [10] depicted in
Figure 2. The data set was obtained from OpenLCA software in ILCD format [4, 17].
Conditioning is an intermediate step in the preparation of logs for the production of ply-
wood [16]. The activity’s requirements are its input flows: hogfuel biomass, electricity
from the grid, heat from the combustion of liquefied petroleum gas (e.g. propane), water,
and the debarked wood itself. Its outcome is its sole output flow, the conditioned logs.
The scope of the activity is a plywood plant in the southeastern United States during
the year 2000. The mix of process energy reported is an averaged result of several
facilities within the scope. To use the data for LCA, each requirement would need to be
linked with an exchange partner that produced it as an outcome. Similarly, this activity’s
outcome could be exchanged with another partner activity that requires the logs as input.
The competency questions presented above can be addressed through the selection
of exchange partners. A database can be constructed that accepts an activity specifica-
tion and a spatiotemporal scope and returns a list of exchange partners. The logistical
requirements associated with performing the exchange are implicit in the spatiotempo-
ral scopes of the partners. Differences in emissions resulting from the same activity in
a different scopes can be inferred through the differences in exchange partners.
Fig. 2. An example to populate LCA Activity pattern using Conditioning Log Activity.
5 Conclusions and Further Work
This paper proposed a compact ontology to capture the spatiotemporal scope of activi-
ties referred to in LCA inventory models. The pattern enables key competency questions
to be addressed by querying a spatiotemporally-explicit data resource. The pattern can
be used as a bridge between sets of activity and flow definitions and spatial modeling
ontologies. Future work will focus on integrating this ontology with other LCA ontolo-
gies and ontology design patterns (such as [8]) in order to further enhance semantic
interoperability in LCA and improve the reproducibility of published LCA studies.
Acknowledgement: The authors would like to thank Kyle Meisterling, Mike Taptich,
Antonio Medrano, Sarah Cashman, Sara Lafia, Bo Pedersen Weidema, Beatriz Rivela,
Johan Tivander, David E. Meyer, and Gary Berg-Cross for their discussions and con-
structive comments.
6
References
1. Abdalla, A., Hu, Y., Carral, D., Li, N., Janowicz, K.: An ontology design pattern for activity
reasoning. In: Proceedings of the 5th Workshop on Ontology and Semantic Web Patterns,
pp. 1–4. CEUR-WS (2014)
2. Bare, J.C., Pennington, D.W., de Haes, H.A.U.: Life cycle impact assessment sophistication.
The International Journal of Life Cycle Assessment 4(5), 299–306 (1999)
3. Bertin, B., Scuturici, V.M., Risler, E., Pinon, J.M.: A semantic approach to life cycle assess-
ment applied on energy environmental impact data management. In: Proceedings of the 2012
Joint EDBT/ICDT Workshops. pp. 87–94. ACM (2012)
4. Ciroth, A., Winter, S.: Openlca 1.4 overview and first steps. Tech. rep., GreenDelta (2014)
5. Curran, M.A.: Environmental Life-Cycle Assessment. McGraw-Hill Professional Publishing
(Jul 1996)
6. Hauschild, M.: Spatial differentiation in life cycle impact assessment: a decade of method
development to increase the environmental realism of lcia. The International Journal of Life
Cycle Assessment 11, 11–13 (2006)
7. ISO 14044: Environmental management — Life cycle assessment — Requirements and
guidelines. ISO, Geneva, Switzerland (2006)
8. Janowicz, K., Krisnadhi, A.A., Hu, Y., Suh, S., Weidema, B.P., Rivela, B., Tivander, J.,
Meyer, D.E., Berg-Cross, G., Hitzler, P., Ingwersen, W., Kuczenski, B., Vardeman, C., Ju,
Y., Cheatham, M.: A minimal ontology pattern for life cycle assessment data. In: Proceed-
ings of the 6th Workshop on Ontology and Semantic Web Patterns (WOP2015) (2015)
9. Levassur, A., Lesage, P., Margni, M., Deschênes, L., Samson, R.: Considering time in lca:
Dynamic lca and its application to global warming impact assessments. Environmental sci-
ence & technology 44(8), 3169–3174 (2010)
10. National Renewable Energy Laboratory: "Conditioned log, at plywood plant, US SE" [data
file in US Life Cycle Inventory database] (2014), http://www.lcacommons.gov/nre
l/, UUID:1aa8ce2b-0bdb-c124-be38-00002cdd561b
11. Potting, J., Schöpp, W., Blok, K., Hauschild, M.: Site-dependent life-cycle impact assess-
ment of acidification. Journal of Industrial Ecology 2(2), 63–87 (1998), http://dx.doi.o
rg/10.1162/jiec.1998.2.2.63
12. Reap, J., Roman, F., Duncan, S., Bras, B.: A survey of unresolved problems in life cycle
assessment. The International Journal of Life Cycle Assessment 13(5), 374–388 (2008)
13. Rebitzer, G., Ekvall, T., Frischknecht, R., Hunkeler, D., Norris, G., Rydberg, T., Schmidt,
W.P., Suh, S., Weidema, B.P., Pennington, D.W.: Life cycle assessment: Part 1: Framework,
goal and scope definition, inventory analysis, and applications. Environ. Int. 30(5), 701–720
(Jul 2004)
14. Stasinopoulos, P., Compston, P., Newell, B., Jones, H.M.: A system dynamics approach in
lca to account for temporal effects—a consequential energy lci of car body-in-whites. The
international journal of life cycle assessment 17(2), 199–207 (2012)
15. Vardeman, C., Krisnadhi, A.A., Cheatham, M., Janowicz, K., Ferguson, H., Hitzler, P., Buc-
cellato, A.P., Thirunarayan, K., Berg-Cross, G., Hahmann, T.: An ontology design pattern
for material transformation. In: Proceedings of the 5th Workshop on Ontology and Semantic
Web Patterns (WOP2014). pp. 73–77 (2014)
16. Wilson, J.B., Sakimoto, E.T.: Module D: Softwood Plywood Manufacturing. Tech. rep., Con-
sortium for Research on Renewable Industrial Materials (CORRIM), Seattle, WA (2004)
17. Wolf, M.A., Düpmeier, C., Kusche, O.: The international reference life cycle data system
(ILCD) format–basic concepts and implementation of life cycle impact assessment (LCIA)
method data sets. In: Proc. 25th EnviroInfo Conference. Schaker-Verlag (2011)
18. Zhang, Y., Luo, X., Buis, J.J., Sutherland, J.W.: LCA-oriented semantic representation for
the product life cycle. Journal of Cleaner Production 86, 146 – 162 (2015)