355
The computational structure of LCA can be used to aggregate inputs and impacts
of a process, or a mix of processes that are part of a network, at various time scales
depending upon how each process is defined (Heijungs & Suh 2002). The columns
of matrix A can represent a set of unit processes within an establishment to assess
establishment metrics, or they can represent establishments that are interacting
within a network. Further, it is computationally trivial to combine both concepts
into a single matrix model.
For example, consider a power plant in a water basin such that there is interest in
how much water consumption is allocated to the power plant from within the water
basin it resides. Consider two LCA processes: one for constructing the power plant
(a “one-time” occurrence of construction) and an additional process (among possible others) for operating the power plant for an output of 1 kWh (“continuous”
material flows during operation). Constructing the power plant is one functional
unit, and operating the power plant to produce 1 kWh is another functional unit.
There is water consumption associated with constructing the power plant, say 100
million gallons (Mgal), an extensive metric (and “absolute,” see next section). Thus,
before the power plant has begun operation because it has yet to produce electricity
output, its kWh water intensity, an intensive metric, is infinite (i.e., 100 Mgal divided
by 0 kWh). Further, this water consumption might have occurred in another water
basin far away.
Consider that during operation the plant consumes 0.5 gal of water for each kWh
generated, an intensive metric that largely describes the water requirements to cool
the steam cycle of the power plant, and that it generates 20 million kWh each year.
It thus consumes 10 Mgal/year (an extensive real-time metric) during operation
in the water basin in which it resides.
If the power plant operates for 10 years, its total water consumption is 100
Mgal + 10 Mgal/year × 10 years = 200 Mgal, and it generates 200 million kWh, for
10-year water intensity of 200 Mgal/200 million kWh = 1 gal/kWh. Half of the
water was for construction and the other half for operation.
If the power plant operates for 40 years, the average water intensity is
100 Mgal + 10 Mgal/year × 40 years = 500 Mgal, and it generates 800 million kWh,
for a water intensity of 500 Mgal/800 million kWh = 0.625 gal/kWh. Here one-fifth of
the water was for construction (the same 100 Mgal), and four-fifths was for operation.
Thus, the longer the assumed operational lifetime of the power plant, the closer
the water intensity of electricity (gal/kWh) of the entire power plant life cycle
becomes to the operational water intensity. The distinction between the temporal
flows of both inputs and outputs is crucial such that lifetime LCA metrics are not
confused with real-time LCA metrics.
This example demonstrates three points:
1. Environmental impacts can be local, regional/global, or both.
2. Full life cycle intensive metrics approach the real-time (operational) intensive
metrics the longer the establishment operates (or alternatively, the life cycle and
real-time metrics converge by the end of the life cycle).
3. How the definition of the functional unit changes when representing an intensive
versus extensive metric.
13 Metrics
The computational structure of LCA can be used to aggregate inputs and impacts
of a process, or a mix of processes that are part of a network, at various time scales
depending upon how each process is defined (Heijungs & Suh 2002). The columns
of matrix A can represent a set of unit processes within an establishment to assess
establishment metrics, or they can represent establishments that are interacting
within a network. Further, it is computationally trivial to combine both concepts
into a single matrix model.
For example, consider a power plant in a water basin such that there is interest in
how much water consumption is allocated to the power plant from within the water
basin it resides. Consider two LCA processes: one for constructing the power plant
(a “one-time” occurrence of construction) and an additional process (among possible others) for operating the power plant for an output of 1 kWh (“continuous”
material flows during operation). Constructing the power plant is one functional
unit, and operating the power plant to produce 1 kWh is another functional unit.
There is water consumption associated with constructing the power plant, say 100
million gallons (Mgal), an extensive metric (and “absolute,” see next section). Thus,
before the power plant has begun operation because it has yet to produce electricity
output, its kWh water intensity, an intensive metric, is infinite (i.e., 100 Mgal divided
by 0 kWh). Further, this water consumption might have occurred in another water
basin far away.
Consider that during operation the plant consumes 0.5 gal of water for each kWh
generated, an intensive metric that largely describes the water requirements to cool
the steam cycle of the power plant, and that it generates 20 million kWh each year.
It thus consumes 10 Mgal/year (an extensive real-time metric) during operation
in the water basin in which it resides.
If the power plant operates for 10 years, its total water consumption is 100
Mgal + 10 Mgal/year × 10 years = 200 Mgal, and it generates 200 million kWh, for
10-year water intensity of 200 Mgal/200 million kWh = 1 gal/kWh. Half of the
water was for construction and the other half for operation.
If the power plant operates for 40 years, the average water intensity is
100 Mgal + 10 Mgal/year × 40 years = 500 Mgal, and it generates 800 million kWh,
for a water intensity of 500 Mgal/800 million kWh = 0.625 gal/kWh. Here one-fifth of
the water was for construction (the same 100 Mgal), and four-fifths was for operation.
Thus, the longer the assumed operational lifetime of the power plant, the closer
the water intensity of electricity (gal/kWh) of the entire power plant life cycle
becomes to the operational water intensity. The distinction between the temporal
flows of both inputs and outputs is crucial such that lifetime LCA metrics are not
confused with real-time LCA metrics.
This example demonstrates three points:
1. Environmental impacts can be local, regional/global, or both.
2. Full life cycle intensive metrics approach the real-time (operational) intensive
metrics the longer the establishment operates (or alternatively, the life cycle and
real-time metrics converge by the end of the life cycle).
3. How the definition of the functional unit changes when representing an intensive
versus extensive metric.
13 Metrics