search

OpenEnergyPlatform / ontology

Repository for the Open Energy Ontology (OEO)
Creative Commons Zero v1.0 Universal
106 stars 23 forks source link

def of "renewable" needs a rework #741

Closed stap-m closed 3 years ago

stap-m commented 3 years ago

Description of the issue

From PR #739:

Current def: Renewable is an origin of portions of matter that replenish on a human time scale. Def of origin: The origin is a quality that indicates where something comes from (its source).

The use case for has origin renewable usually is to indicate which energy carriers are renewable energy carriers, such as air, photon, rock, water, biomass. The def of renewable fits for biomass, bur not really for the other examples.

Task: Develop a suitable definition for renewable, that fits other renewable energy carriers as well. Question: Is the restriction to portion of matter useful?

State of discussion (21-07-28)

Slides from dev meeting 23: issue741-renewable.pdf

Results so far:

Workflow checklist

I am aware that

stap-m commented 3 years ago

From a modellers perspective it would be useful that energy could also have an origin directly, athough we defined it as a quality of matter.

l-emele commented 3 years ago

The relation has origin currently has the domain energy or 'portion of matter'. After the latest verbal discussion and some thoughs on hydro-related questions I am asking myself, whether we should limit the domain to energy or 'fuel'.

It is irrelevant whether rock renews itself or not, what is really renewed is the thermal energy carried by rock. And for the hydro-related classes attributing water the renewable origin causes some at least to me some trouble when thinking of pumped-storage hydroelectricity.

sfluegel05 commented 3 years ago

If the term renewable is used for energies, we should make it applicable to energies (and do the same for other origins). For a new definition of renewable this might be helpful source:

Renewable energy is energy that is generated from natural processes that are continuously replenished.

This emphasizes that the origin of something is related to the process which generated it. For renewable this could mean Renewable is an origin of a portion of mater or energy that is the output of a natural process that is continuously replenished.

The relation has origin currently has the domain energy or 'portion of matter'. After the latest verbal discussion and some thoughs on hydro-related questions I am asking myself, whether we should limit the domain to energy or 'fuel'.

I agree that renewable does not really work for rock or water. But other origins might still be useful for non-fuels. E.g. biogenic for certain instances of air pollutant or anthropogenic for artificial object (which would mean that we extend the range of has origin to material entity). That doesn't mean we need to add origins to everything, but I think it would be a good option to have.

stap-m commented 3 years ago

@l-emele @han-f do you think it could be relevant to declare the origin of e.g. emissions? I think the OEO stays more flexible when allowing the relation has origin to any portion of matter and energies.

The class origin itself isn't resticted btw. Maybe this should be adjusted alongside...

han-f commented 3 years ago

Without thinking too much about it, personally it seems to come naturally to have the possibility to attach an origin to many things. Would we then be able to declare an emission to be originating from electricity production for example, or from road transportation (a kind of sectoral origin)?

If yes and we do not currently have this implemented yet in another way: I would definitely be in favour to also declare origins of emissions as this is helpful in reporting contexts. For example, European Member States need to report emission projections by sector according to table 1a in Annex XXXV: https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32020R1208&from=EN Thus they also utilise models and tools that allow for such a depiction. This then ties in with the sector division common reporting format.

Sidenote: In the linked regulation the tables are called ‘common reporting table’, or ‘CRT’ which means a table for information on anthropogenic greenhouse gas emissions by sources and removals by sinks included in AnnexII to Decision 24/CP19 of the Conference of the Parties to the United Nations Framework Convention on Climate Change (UNFCCC) (Decision 24/CP19);, and I am not 100 % sure they match spot on with the common reporting format, but they look very much so and may slightly differ in their depth.

l-emele commented 3 years ago

Without thinking too much about it, personally it seems to come naturally to have the possibility to attach an origin to many things. Would we then be able to declare an emission to be originating from electricity production for example, or from road transportation (a kind of sectoral origin)?

We already have a relation describing this, which is has output resp. output of.

I understand the concept of origin differently: Its the source of something without being the output of an other process we describe in the ontology.

han-f commented 3 years ago

Ah ok, that both makes sense.

stap-m commented 3 years ago

But we have the subclass of origin anthropogenic, which is quite suitable for emissions. I don't see a conflict with the output of here.

Can we agree to stay with the current def of origin? The origin is a quality that indicates where something comes from (its source).

l-emele commented 3 years ago

@han-f : What exactly do you mean by emission, btw? In our current structure, we distinguish between emission (a process), greenhouse gas (a portion of matter) and emission value (process attribute). What you want to describe is probably the origin of the gases and not of the emission process. For some of the gases we already have included anthropogenic origins namely for nitrogen trifluoride, sulphur hexafluoride, hydrofluorocarbon and perfluorocarbon.

han-f commented 3 years ago

@han-f : What exactly do you mean by emission, btw? In our current structure, we distinguish between emission (a process), greenhouse gas (a portion of matter) and emission value (process attribute). What you want to describe is probably the origin of the gases and not of the emission process. For some of the gases we already have included anthropogenic origins namely for nitrogen trifluoride, sulphur hexafluoride, hydrofluorocarbon and perfluorocarbon.

Thanks for this explanation and yes, taking this into account the origin would refer to the gases.

stap-m commented 3 years ago

I understand the concept of origin differently: Its the source of something without being the output of an other process we describe in the ontology.

What about 1) Origin is a quality that indicates where a portion of matter or energy comes from (its source). 2) Adding an editor note to origin like: Origin is not used to sedcribe the output of a process. Here, the relation "has physical output" is recommended.

han-f commented 3 years ago

I like your suggestion of an editor note - repeated here with typo corrected and a slightly adapted second sentence: Origin is not used to describe the output of a process. In such a case, we recommend to use the relation "has physical output" .

stap-m commented 3 years ago

@l-emele do you agree, too?

l-emele commented 3 years ago

I agree with the definition proposal and the editor note.

stap-m commented 3 years ago

Great. What about smth like this for renewable: Renewable is an origin of an energy carrier or energy that indicates that the energy carrier replenishes on a human time scale or isn't consumed when the respective energy is used. ?

l-emele commented 3 years ago

I think, I agree too this definition of renewable. But I still have a small doubt because of the water problem.

The class water has currently the axiom has origin some renewable but for the subclass pumped water this axiom is not necessarily true. I cannot imagine a use case where water, that was pumped into a reservoir using electricity generated by a coal power plant would be considered as renewable.

My impression is, solving the water problem first might lead to an even better definition of renewable.

stap-m commented 3 years ago

Methane is probably a similar case. Both can have origin renewable, but not necessarily. My first idea is to create parallel classes methane and renewable methane / pumped water and renewable pumped water? But in the case of water, the origin is inherited. Thus, this won't work. Any good ideas @sfluegel05 ?

sfluegel05 commented 3 years ago

I guess the easiest solution would be to not use any axioms has origin some renewable, if some instances of the class are not renewable. So we would only use renewable for cases which are unambiguous, like biofuel.

Another thought: Is renewable really an origin? An origin indicates where something comes from, but the definition of renewable doesn't make a statement about the past, but about the future: the energy carrier replenishes on a human time scale.

sfluegel05 commented 3 years ago

Also, I'm afraid we have to rethink the origin-energy-relation. Our current suggestions all include that an energy can have an origin. However, the relation has origin is a subproperty of has quality, which is defined as a relation between an independent continuant (the bearer) and a quality, in which the quality specifically depends on the bearer for its existence Unfortunately, energy (just like origin) is a quality and not an independent continuant. We could

  1. make energy an independent continuant (which would probably be a misclassification, since energy depends on an energy carrier and thus is not independent)
  2. apply origin only to material entities, i.e. the energy carriers
l-emele commented 3 years ago

We could

1. make `energy` an independent continuant (which would probably be a misclassification, since energy depends on an energy carrier and thus is not independent)

2. apply `origin` only to material entities, i.e. the energy carriers

Of these two options I would prefer the second one. A misclassification of the very important class as energy should be avoided.

l-emele commented 3 years ago

I have an idea for a potential solution:

With this "trick" of limiting renewable first to energies only, but then going via the renewable energy carrier disposition we still can define a renewable energy carrier and a renewable fuel. Or did I miss something?

l-emele commented 3 years ago

Additional minor thing: In any case we should add the axiom renewable disjoint with fossil.

sfluegel05 commented 3 years ago

This sounds like a viable solution. Only a minor change: I would define a renewable energy carrier disposition as an energy carrier disposition of an material entity to contain renewable energy. to contain instead of that contains, because the material entity doesn't have to carry the energy yet in order to have an energy carrier disposition (other energy carrier disposition-subclasses also use that, we should change that as well).

Let's apply it to the class water to test this solution.

  1. Is hydro energy renewable?

    • Hydro energy is kinetic energy of moving liquid water which can result directly from its potential energy.
    • this replenishes on a human time scale (a reservoir can be refilled and water can be released again)
    • hydro energy should get the axiom has origin some renewable

  2. Does water have the disposition to carry renewable energy?

    • hydro energy uses some water flow
    • water flow has participant some liquid water
    • this connects hydro energy to liquid water, implying that liquid water can carry renewable energy
    • liquid water should get the axiom has disposition some renewable energy carrier disposition and will be classified as a renewable energy carrier

The "water problem" (https://github.com/OpenEnergyPlatform/ontology/issues/741#issuecomment-866064915) would be solved by not saying that "this pumped water is renewable", but instead saying that "this pumped water can potentially carry renewable energy". The problem is now: is hydro energy renewable? It could still be the result of potential energy of water that has been pumped upwards by a coal power plant.

The other problem we still need to solve: How can we connect origin to energy axiomatically?

PS: I attached the slides from dev meeting 23 to the issue header.

sfluegel05 commented 3 years ago

in dev meeting 23 this was also brought into the discussion:

why did we not decide on a process for origin? reasoning:

  • the energy ending up in an energy carrier or being available is the output of a process
  • Renewable: the process needs to be repeatable and does only require renewable inputs (!circular logic!) / and is fed with energy from a replenishable (on a human timescale) source (stealing from the current definition)
  • none of the above is good - discuss again
  • LE, MS or HF can prepare (participants of issue)
l-emele commented 3 years ago

Regarding the "water problem" I would argue that hydro energy in general is not renewable. However, we could differentiate between subclasses of hydro energy:

We do not have a concept renewable electrical energy yet but it could be defined as an equivalent class electrical energy and has origin some renewable (or even electrical energy and has origin only renewable?