search

OpenEnergyPlatform / ontology

Repository for the Open Energy Ontology (OEO)
Creative Commons Zero v1.0 Universal
106 stars 22 forks source link

potentials should get included in the oeo #481

Closed akleinau closed 3 years ago

akleinau commented 4 years ago

Description of the issue

Potentials are described in the oeo scenario factsheets but not in the ontology and should be included: theoretical, technical, economic, developable potential

Current state of discussion (11/11/20)

There are two fundamentally different ideas:

1. Implement potential as a disposition.

For this, we have the following classes and relations:

2. Implement has potential as an object property.

Why should we chose the disposition- or the object-property-solution?

Workflow checklist

I am aware that

0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q commented 3 years ago

Do you have an idea for a better wording?

Reading this thread again, I think that the whole disposition direction is wrong. To say that the potential of oil production is an attribute of the oil strikes me as odd. For solar/wind/geothermal, this relation breaks down entirely, since there's no material. For CO2 storage, it's the absence of material that counts. We are not talking about the potentials of oil, wind energy, or CO2 storage, but about the potentials to produce oil (as in primary energy carrier), to produce electricity from wind, and to store CO2. So these are processes that can have quantity values (Joules, Watts, tonnes) attached to them through a bunch of relations has (theoretical|technical|economic|developable|sustainable|…) potential.

stap-m commented 3 years ago

Reading this thread again, I think that the whole disposition direction is wrong. To say that the potential of oil production is an attribute of the oil strikes me as odd. For solar/wind/geothermal, this relation breaks down entirely, since there's no material. For CO2 storage, it's the absence of material that counts.

Are you referring to the fact, that a disposition needs an independent continuant as bearer? That's a valid point. Btw, quantity values are generically dependent continuants and thus also dependent on an independent continuant as bearer, not on processes... although the def of has quantity value is vague here. Did we do that intentionally to include processes or is that an accident? 👀

jannahastings commented 3 years ago

The relationship between quantity values as generically dependent continuants and the things they are measures of or counts of is is about , not dependence. They can be is about any sort of entity.

They depend on a different sort of thing -- a kind of bearer which is generic, hence the generic in the name generically dependent continuant. This typically means how is the value encoded or stored.

sfluegel05 commented 3 years ago

We are not talking about the potentials of oil, wind energy, or CO2 storage, but about the potentials to produce oil (as in primary energy carrier), to produce electricity from wind, and to store CO2. So these are processes that can have quantity values (Joules, Watts, tonnes) attached to them through a bunch of relations has (theoretical|technical|economic|developable|sustainable|…) potential.

This is an interesting view on the topic. In accordance with our "quantity value-scheme" (https://github.com/OpenEnergyPlatform/ontology/issues/434#issuecomment-648837549) which was the reason we classified potentials as dispositions in the first place, we could say that the process is the entity in reality and the has ... potential relations are subrelations of has quantity value.

sfluegel05 commented 3 years ago

For definitions we could use the term output, although I'm not sure if this makes the distinction towards has output clear enough. The definitions would look like this:

has potential: A relation between a process and the upper limit of its usable output. has technological potential: A relation between a process and the upper limit of its usable output derived from a theoretical potential, taking account of the annual efficiency of the respective conversion technology and the additional restrictions regarding the area that is realistically available for output generation.

l-emele commented 3 years ago

It's not necessary an input. The CO2 storage potential is more an input than an output.

0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q commented 3 years ago

The CO2 storage potential is more an input than an output.

The CO2 that is stored is an output of the process of storing CO2, so has potential would work. But there is neither "conversion" nor "generation", and the CO2 isn't "usable", so has theoretical potential wouldn't.

🤔

This is more free association than a concrete suggestion:
Would it be useful to define a magnitude of a process, that is specific to the process? For energy transformation that would be power, for consumption that would be the amount of what is consumed per unit time, etc. The potentials would then be the upper limits to the magnitudes. This would encapsulate the specifics of the processes and allow for a orthogonal definition of potentials, which would be handy for future-proofing for processes to be added (CO2 storage isn't in there yet. Water withdrawal becomes popular with energy models looking into sustainability aspects.)

Not sure how to implement this cleanly in terms of classes and attributes, though.

l-emele commented 3 years ago

The CO2 storage potential is the upper limit of CO2 input. The process where the CO2 comes from is irrelevant.

0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q commented 3 years ago

The CO2 storage potential is the upper limit of CO2 input.

Input into what?

l-emele commented 3 years ago

Of the thing the potential is given for, e.g. the geological formations in Germany.

0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q commented 3 years ago

We moved past that in the discussion. We're at potentials of processes, therefore: "[t]he CO2 that is stored is an output of the process of storing CO2".

It still ends up in the geological formation, but basing the definition on the processes allows for a generalised definition of potentials (maybe).

l-emele commented 3 years ago

I disagree. In my understanding the output of the CO2 storage process would be the (unintentional) leakage of CO2 from the storage site back into the atmosphere.

l-emele commented 3 years ago

this issue has 42 comments. Maybe it is a good idea to define an upper limit like 30 comments, after which an issue should be discussed in a dev meeting as it got too complex?

We could also think about handing over this topic to our steering committee as this is the original purpose of the OEO-SC.

stap-m commented 3 years ago

Or discuss it at the next dev-meeting first, as originally proposed. Afterwards we can still forward it to the SC, if needed.

I'd propose to close this issue now and reopen it at the next meeting.

0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q commented 3 years ago

I disagree. In my understanding the output of the CO2 storage process would be the (unintentional) leakage of CO2 from the storage site back into the atmosphere.

  1. I think this discussion is moot in the context of this issue, since the more specific potentials don't work for CCS for other reasons.

  2. As a general point:
    That depends on how the process boundaries are defined, and that's completely arbitrary. Our model (REMIND) uses four different processes for modelling CCS (in addition to energy conversion processes with carbon capture):

        ccscomp         "compression of co2"
    ccspipe         "transportation of co2"
    ccsinje         "injection of co2"
    ccsmoni         "monitoring of co2"

    Fugitive CO2 emissions in that setup result from the ccsmoni process.

0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q commented 3 years ago

We could also think about handing over this topic to our steering committee as this is the original purpose of the OEO-SC.

Or discuss it at the next dev-meeting first, as originally proposed. Afterwards we can still forward it to the SC, if needed.

I'd propose to close this issue now and reopen it at the next meeting.

What's the point here? I didn't have the impression that discussing an issue with lots of people who haven't read through the issue before is particularly effective or efficient.

I don't think that we are talking past one another and struggle with ambiguities that are cleared up more easily in direct communication. In my view we had a couple of proposals that all were rejected because they didn't give a proper definition of potentials that fulfils all requirements. Having a dozen people chew the fat about it will most likely not yield a solution either.

As for the steering committee, what would be their assignment here? There's no decision to be made or dispute to settle. So they should come up with a working solution, because we couldn't? If so, more power to them. But I think that's not the case.

sfluegel05 commented 3 years ago

I tried to summarize the discussion so far in the first comment. Maybe this helps for a discussion in the dev-meeting.

l-emele commented 3 years ago

Definition proposals from OEO dev meeting 11:

0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q commented 3 years ago

I can't help but notice that no @OpenEnergyPlatform/oeo-general-expert-formal-ontology was assigned to this issue, and I think one should be.

  • A flow potential is a quantity value that describes the upper limit of an input or output value of a process in an n-dimensional region per time unit. For example the wind flow potential of Germany is the amount of energy available to wind power plants in Germany.
  • A stock potential is a quantity value that describes the upper limit of a stock value of a source or sink in an n-dimensional region. An example is the coal potential of Germany that is the amount of coal available in the soil in Germany.
l-emele commented 3 years ago 
* What's with the _n-dimensional region_?  There's no such thing in the ontology.  Should it be?
  I guess it's supposed to be `spatial region`.

Yes, it should be a spatial region. I forgot yesterday the precise label of that region class.

Is there any instance in which potentials are not defined over an area, but either for a point, a line, or a volume? Fossil resources and CO2-storage could be defined for different strata, but is anybody doing it? And if so, wouldn't it be more straight-forward to include this in an annotation (or some such) to the potential, like all those strings attached to technological potential and so on?

Spatial region has currently the subclasses zero-dimensional region (a point), two-dimensional region (an area) and three-dimensional region (a volume). This is why I called this yesterday n-dimensional, A one-dimensional region covering a line is currently missing, but if needed we can definitely create one. But that is topic for a different issue.

* This is a differential quotient, not an average, so it should read "per unit time", not "per time unit".

Right.

* "Wind flow potential"?  I'd say `wind energy potential`.

That is about the label of potential subclasses of potential. There are more than one way to find a proper label. One is to derive it from the parent label, that would be the wind flow potential, the other way is to use a label that is close a term that is used in the domain like wind energy potential. Important: the two options are not exclusive, we can e.g. label the class wind flow potential and additionally give the class the alternative term wind energy potential. One example in our ontology is the class wind rotor which can also be found by its alternative term wind turbine.

  • A stock potential is a quantity value that describes the upper limit of a stock value of a source or sink in an n-dimensional region. An example is the coal potential of Germany that is the amount of coal available in the soil in Germany.
* Coal is by definition not in the soil.  "… available for production in Germany" will do.

* This presupposes that the stock value of "the amount of coal available […] in Germany" is included in the ontology, in order to attach the quantity value, right?

To me, this term available for production would imply something like a technological potential, but the example is about a potential in general. What about ... amount of coal available below the surface of Germany.?

stap-m commented 3 years ago

I'd exclude the example from the def. There is an annotation "example of usage", where the example can be included. And let's remove the terms "value" from the def. I adjusted as follows: A flow potential is a quantity value that describes the upper limit of an input or output of a process in a spatial region of reference per unit time. A stock potential is a quantity value that describes the upper limit of a stock of a source or sink in a spatial region of reference.

And I'd rather not implement a class wind energy/flow potential. We have relations for that.

l-emele commented 3 years ago

I'd exclude the example from the def. There is an annotation "example of usage", where the example can be included. And let's remove the terms "value" from the def. I adjusted as follows: A flow potential is a quantity value that describes the upper limit of an input or output of a process in a spatial region of reference per unit time. A stock potential is a quantity value that describes the upper limit of a stock of a source or sink in a spatial region of reference.

I agree.

And I'd rather not implement a class wind energy/flow potential. We have relations for that.

Good point. Procedural suggestion: We implement the potential classes and then close this issue.

For the application of these classes we should create a separate issue. Also we should keep in mind that we decided yesterday that concrete realisations of potentials (like wind energy potential in Germany) will be individuals. I forgot that, too, when writing my last comment.

Vera-IER commented 3 years ago

Procedural suggestion: We implement the potential classes and then close this issue.

Do you mean by potential classes only the flow potential and the stock potential or also their subclasses like theoretical/technical/economic flow potential and so on?

l-emele commented 3 years ago

Procedural suggestion: We implement the potential classes and then close this issue.

Do you mean by potential classes only the flow potential and the stock potential or also their subclasses like theoretical/technical/economic flow potential and so on?

Also these subclasses but not things like wind energy potential.

Vera-IER commented 3 years ago

I summed up the subclasses for flow potential and slighly addjusted their definitions: flow potential: A flow potential is a quantity value that describes the upper limit of an input or output of a process in a spatial region of reference per unit time. subclasses: theoretical flow potential: A theoretical flow potential is a type of flow potential that identifies the physical upper limit of an input or output of a process in a spatial region of reference per unit time. technological flow potential: A technological flow potential is a type of a flow potential derived from a theoretical flow potential, taking account of the annual efficiency of the respective conversion technology and the additional restrictions regarding the area that is realistically available for energy generation. economic flow potential: An economic flow potential is a type of flow potential that identifies the proportion of the technological potential that can be utilized economically (based on economic boundary conditions). developable flow potential: A developable flow potential is a type of flow potential that describes the fraction of the economic potential that can be developed under realistic conditions (regulations, environmental and social restrictions). sustainable flow potential: A sustainable flow potential is a type of flow potential that takes into account all aspects of sustainability, which usually requires careful consideration and evaluation of different ecological and socio-economic aspects. The differentiation of the sustainable potential is blurred, since ecological aspects may already have been considered for the technological or economic potential, depending on the author.

If you agree to the definitions, we would need to do the same for stock potential, right? Then implement them and open a new issue for the relations sounds good to me.

0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q commented 3 years ago

Spatial region has currently the subclasses zero-dimensional region (a point), two-dimensional region (an area) and three-dimensional region (a volume). This is why I called this yesterday n-dimensional, A one-dimensional region covering a line is currently missing, but if needed we can definitely create one. But that is topic for a different issue.

Nope, I didn't "see things." It's there. https://github.com/OpenEnergyPlatform/ontology/blob/062f910c32821d0e1d075dd524b23e8399a4f49c/src/scripts/iao/iao.owl#L2529-L2539

But my point is that potentials will only be defined over areas, never over points, lines, or volumes. At least I can't think of any useful application for those. So the definitions would be more specific and less ambiguous if they were defined over an area instead of any spatial region.

flow potential: A flow potential is a quantity value that describes the upper limit of an input or output of a process in a two-dimensional spatial region of reference per unit time. stock potential: A stock potential is a quantity value that describes the upper limit of a stock of a source or sink in a two-dimensional spatial region of reference.

And could we please include the alternative terms point, line, area, and volume for the spatial regions? What was good enough for Euclid should also be good enough for BFO. ;)

flow potential: A flow potential is a quantity value that describes the upper limit of an input or output of a process in a reference area per unit time. stock potential: A stock potential is a quantity value that describes the upper limit of a stock of a source or sink in a reference area.

l-emele commented 3 years ago

EDIT: Just two counter examples: You can definitely give a wind energy potential for Germany in the atmosphere up to a height of 200 m. So this is a 3-dimensional region. Also the worldwide coal potential has a 3-dimensional reference region as the Earth is a 3-dimensional object and the coal potential is definitely not limited to the coal lying on the Earth's surface.

The consensus in the dev meeting was n-dimensional, not 2-dimensional. Let's stick with that consensus.

0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q commented 3 years ago

You can definitely give a wind energy potential for Germany in the atmosphere up to a height of 200 m. So this is a 3-dimensional region. Also the worldwide coal potential has a 3-dimensional reference region as the Earth is a 3-dimensional object and the coal potential is definitely not limited to the coal lying on the Earth's surface.

Sure, one can. But nobody does – as far as I know. But who cares about usability …

l-emele commented 3 years ago

Just one further example, this time from literature.

Global geothermal technical potential is comparable to global primary energy supply in 2008. For electricity generation, the technical potential of geothermal energy is estimated to be between 118 EJ/yr (to 3 km depth) and 1,109 EJ/yr (to 10 km depth).

Source: Executive Summary of Chapter 4 of the IPCC Special Report on Renewable Energy Sources and Climate Change Mitigation, page 404.

The IPCC provides here a geothermal technical potential for spherical shells which are 3-dimensional regions.

stap-m commented 3 years ago

Then let's stick with spatial region and implement.