Open 0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q opened 4 years ago
We also just converted some of our model output data in this IAMC format for another project. So for the LOD GEOSS prototyp I can also give the intended output data in this format (but we added some variables).
I think all these variables don't need to be implemented in the ontology individually. For exapmle for "Primary energy|coal/gas/oil" I would use the term primary energy
(which is not included yet) and assemble it with the term coal
,gas
or oil,
which is implemented under portion of matter
.
We also just converted some of our model output data in this IAMC format for another project. So for the LOD GEOSS prototyp I can also give the intended output data in this format (but we added some variables).
So you have seen the light :) What data did you report?
I think all these variables don't need to be implemented in the ontology individually.
Gosh, I hope this is true. The potentials OpenEnergyPlatform/ontology#481 didn't raise my hopes in this regards …
For exapmle for "Primary energy|coal/gas/oil" I would use the term
primary energy
(which is not included yet) and assemble it with the termcoal
,gas
oroil,
which is implemented underportion of matter
.
Do you know how this ("assembling") works? So far I haven't been able to get any answers regarding the use of the ontology.
I wonder how unambiguous primary energy
∩ coal
is. I think it would be contrary to the idea of an ontology if it were just an undocumented convention for that to mean "coal primary energy consumption". So in my view there has to be some "ontological" mechanism to achieve this, else it's lots of new classes – an probably a script to spell them out.
What data did you report?
We will report the final energy consumptions, the capacities, the secondary energy|electricity|..., and the emissions.
Do you know how this ("assembling") works?
I think this is done with the relations between classes of the ontology. I understood that a user of the ontology can add relations for their specific needs in between classes, even if they aren't implemented yet. But someone from @OpenEnergyPlatform/oeo-general-expert-formal-ontology can propably answer this question better than me.
I wonder how unambiguous
primary energy
∩coal
is.
You would not use an intersection for this -- logically, the intersection of primary energy
and coal
is empty since nothing is an instance of both of these classes. Rather, you would model an explicit relationship or other logical axiom that formally captured the meaning of the "defined class" (what Vera calls assembled). Assuming that you want to define a class primary energy consumption from coal
, then I would imagine that has input
might be a good candidate relationship, in which case the formal definition for primary energy consumption from coal
, a subclass of primary energy consumption
, would be 'primary energy consumption' and 'has input' some coal
So you are describing adding a new class to the ontology. As I understood Vera, her approach was to link data items (e.g. our variable Primary Energy|Coal
) to more than one class from the ontology (what she called "assembling").
I understood that a user of the ontology can add relations for their specific needs in between classes
I read that as
but maybe I got that wrong.
In any case these are two different approaches as to how the ontology is used. Are both viable?
And again, a real-world example of some data (any data, really) that is annotated by an ontology (any ontology, really) would be very handy.
Yes, there are two approaches. Technically these two approaches are called "pre-composition" (adding new classes to the ontology for combinations of existing classes) and "post-composition" (using combinations of ontology classes on-the-fly in annotations). You can read a nice description of the difference in this paper: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2847714/.
At the stage where the OEO is currently at in its development (it is not so large an ontology) and tool support available for our users (we do not have a budget to develop custom tools to support all use cases), I would recommend that we add classes to the OEO for all the terms needed for annotations (i.e. using the pre-composed approach) because this reduces the burden on the users for adoption.
Thank you for the clarification! I thought both options can be easily realised.
So in this example we would have the class primary energy consumption
and some relationships like
primary energy consumption has input some coal
primary energy consumption has input some gas
and so on. So I understand that these relationships are subclasses of primary enery consumption
and I can choose if I want to use a relation at all or I can choose just one of the existing relations as a user of the ontology? Is this correct?
Thank you for the clarification! I thought both options can be easily realised.
Me too. I always thought we should better not flush the OEO with all these subclasses if users can use relations. But I understand that for users it might be more comfortable with the pre-composed approach at this state...
primary energy consumption has input some coal
we have the brand new has physical input
now ;) The generic relation primary energy consumption has physical input some primary energy carrier
should be implemented. Own issue for that?
@0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q the first two mandatory entries are "Final Energy" and "Primary Energy". According to the def it is similar to final energy consumption
and primary energy consumption
respectively. From your point of view, what is needed to satisfy the IAMC template requirements?
final energy
and primary energy
?final energy according to IAMC
as subclass of final energy consumption
, specifying the definition (see above)?final energy consumption
: Final energy consumption is the consumption of energy delivered to and consumed by end users.
and
Final Energy
: total final energy consumption by all end-use sectors and all fuels, excluding transmission/distribution losses
match up and can be used as is.
There's a material difference between the definitions of primary energy consumption
in the ontology and Primary Energy
, as used in the IAMC, in that the former excludes non-energy use and the latter includes it. So that is a no-go.
primary energy consumtion
: Primary energy consumption is the total consumption of energy in a spatial region excluding the non-energetic use of fuels.
Primary Energy
: total primary energy consumption (direct equivalent)
So maybe we should distinguish between PE excluding non-energy use (PE according to Eurostat) and PE including non-energy use (PE according to UNSTATS).
Provided that we know models that actually use the Eurostat definition @OpenEnergyPlatform/oeo-domain-expert-energy-modelling because to me that looks like a bookkeeper's definition, not like that of a modeller.
We use Eurostat and AG Energiebilanzen to calibrate the TIMES PanEU model. These statistics report energy consumption excluding non-energy uses. So I would vote for distinguishing between those two reporting options. The same problem can arise when defining Final energy|Industry. We define it excluding non-energy uses, some models define it including the non energetic fuel uses.
How do we proceed with this long meta issue? I suggest adding an additional column in the long table indicating what is there and what is missing.
Out of the 29 mandatory variables I would judge two to be implemented in usable form.
Variable | Definition | OEO IRI | Notes | |
---|---|---|---|---|
✓ | Final Energy | total final energy consumption by all end-use sectors and all fuels, excluding transmission/distribution losses | final energy consumption | |
✓ | Population | total population | population | |
✗ | Primary Energy | total primary energy consumption (direct equivalent) | gross inland energy consumption | OEO is ambiguous as to whether direct equivalent/physical energy content/partial substitution are used to compute primary energy of non-fuels. See https://www.iea.org/statistics-questionnaires-faq. It comes down to whether 1 EJ of electricity from e.g. wind is credited with 1 EJ primary energy wind, 1.7 EJ primary energy wind (using the 60 % efficiency of the wind turbine), or 2.5 EJ primary energy wind (using the 40 % efficiency of the conventional power plant fleet that would have to replace the wind turbines). |
✗ | Primary Energy|Biomass | primary energy consumption of purpose-grown bioenergy crops, crop and forestry residue bioenergy, municipal solid waste bioenergy, traditional biomass | ibid. | |
✗ | Primary Energy|Fossil | coal, gas, conventional and unconventional oil primary energy consumption | ibid. | |
✗ | Primary Energy|Fossil|w/ CCS | coal, gas, conventional and unconventional oil primary energy consumption used in combination with CCS | ibid. | |
✗ | Primary Energy|Fossil|w/o CCS | coal, gas, conventional and unconventional oil primary energy consumption without CCS | ibid. | |
✗ | Primary Energy|Coal | coal primary energy consumption | ibid. | |
✗ | Primary Energy|Gas | gas primary energy consumption | ibid. | |
✗ | Primary Energy|Nuclear | nuclear primary energy consumption (direct equivalent, includes electricity, heat and hydrogen production from nuclear energy) | ibid. | |
✗ | Primary Energy|Oil | conventional & unconventional oil primary energy consumption | ibid. | |
✗ | Primary Energy|Non-Biomass Renewables | non-biomass renewable primary energy consumption (direct equivalent, includes hydro electricity, wind electricity, geothermal electricity and heat, solar electricity and heat and hydrogen, ocean energy) | ibid. | |
✗ | Secondary Energy|Electricity | total net electricity production | secondary energy production has no subdivision by energy carrier classes | |
✗ | Secondary Energy|Electricity|Biomass | net electricity production from municipal solid waste, purpose-grown biomass, crop residues, forest industry waste, biogas | ibid. | |
✗ | Secondary Energy|Electricity|Coal | net electricity production from coal | ibid. | |
✗ | Secondary Energy|Electricity|Gas | net electricity production from natural gas | ibid. | |
✗ | Secondary Energy|Electricity|Oil | net electricity production from refined liquid oil products | ibid. | |
✗ | Secondary Energy|Electricity|Nuclear | net electricity production from nuclear energy | ibid. | |
✗ | Secondary Energy|Electricity|Non-Biomass Renewables | net electricity production from hydro, wind, solar, geothermal, ocean, and other renewable sources (excluding bioenergy). This is a summary category for all the non-biomass renewables. | ibid. | |
✗ | Secondary Energy|Liquids | total production of refined liquid fuels from all energy sources (incl. oil products, synthetic fossil fuels from gas and coal, biofuels) | ibid. | |
✗ | Emissions|CO2 | total CO2 emissions (not including CCS) | carbon dioxide equivalent quantity 'has quantity value' some 'emission quantity value' (as in “MtCO2eq”) but has no reference to time (as in “per year”) and is not specific to CO2 either |
|
✗ | Emissions|CO2|Energy and Industrial Processes | CO2 emissions from energy use on supply and demand side (IPCC category 1A, 1B) and from industrial processes (IPCC categories 2A, B, C, E) | ibid. | |
✗ | Emissions|CO2|Industrial Processes | CO2 emissions from industrial processes (IPCC categories 2A, B, C, E) | ibid. | |
✗ | Emissions|CO2|Energy | CO2 emissions from energy use on supply and demand side (IPCC category 1A, 1B) | ibid. | |
✗ | Carbon Sequestration|CCS | total carbon dioxide emissions captured and stored in geological deposits (e.g. in depleted oil and gas fields, unmined coal seams, saline aquifers) and the deep ocean, stored amounts should be reported as positive numbers | carbon capture and storage | a process without any attribute denoting magnitude (as in MtCO2/year) |
✗ | Carbon Sequestration|CCS|Biomass | total carbon dioxide emissions captured from bioenergy use and stored in geological deposits (e.g. in depleted oil and gas fields, unmined coal seams, saline aquifers) and the deep ocean, stored amounts should be reported as positive numbers | ibid. | |
✗ | Carbon Sequestration|CCS|Fossil | total carbon dioxide emissions captured from fossil fuel use and stored in geological deposits (e.g. in depleted oil and gas fields, unmined coal seams, saline aquifers) and the deep ocean, stored amounts should be reported as positive numbers | ibid. | |
✗ | GDP|PPP | GDP at purchasing power parity | gross domestic product | both GDP and PPP are subclasses of economic value ; not sure how to combine them |
✗ | Price|Carbon | price of carbon (for regional aggregrates the weighted price of carbon by subregion should be used) | emission certificate price | A carbon price might be implemented as a tax instead. There is carbon tax, but the IAMC template is agnostic towards the implementation of the carbon price. |
OEO is ambiguous as to whether direct equivalent/physical energy content/partial substitution are used to compute primary energy of non-fuels. See https://www.iea.org/statistics-questionnaires-faq. It comes down to whether 1 EJ of electricity from e.g. wind is credited with 1 EJ primary energy wind, 1.7 EJ primary energy wind (using the 60 % efficiency of the wind turbine), or 2.5 EJ primary energy wind (using the 40 % efficiency of the conventional power plant fleet that would have to replace the wind turbines).
Seems you need a special type of primary energy consumption. We can define a subclass for that.
Regarding the combined classes: We probably cannot include every combined concept, but there is also option of post composition. For example Emission|CO2 can be expressed as 'emission quantity value' 'quantity value of' some ('greenhouse gas emission value' 'process attribute of' some ('greenhouse gas emission' 'has participant' some 'carbon dioxide'))
. I know, that is not very nice, but at least it is possible.
I know of post composition, but not about it. Has its technical realisation been discussed in one of the OEO dev meetings?
What is the state of this issue? Last comment is more than 1½ years. Is it still planned to align the IAMC template and the OEO? If not, I suggest to close this issue.
EDIT: If this the IAMC alignment is still on the agenda, the IAMC terms are a good candicates for the planned OEO composed module.
A majority of the composed concepts can be implemented here. Thus, I transferred the issue.
Description
This originated from the OEO Dev Meeting 11 and inteded to provide an overview of the variables from the "IAMC-Template" (which isn't concretely defined) that should be implemented in the ontology for use with Integrated Assessment Models. As a basis I used the submission template to the scenario data base underlying the IPCC Special Report on Global Warming of 1.5°C. This might be extended with variables from the submission template of the Sixth Assessment Report once it is finalised, but I do not expect much development of the variables in the energy domain.
The submission template covers 540 variables which are subdivided in four tiers (mandatory, high priority, medium priority, other). Variables of the first two tiers (mandatory and a sizeable share of high priority) should be included in the ontology for use within the LOD-GEOSS project.
Variable definitions are quite orthogonal, with different subdivisions repeating over many top-level variables within specific groups.
Mandatory Variables (29 Variables)
HIgh Priority Variables (173 Variables)
Medium Priority (235 Variables)
Other (103 Variables)