include subclasses of energy transformation

Vera-IER commented 4 years ago

Description of the issue

We need some subclasses of "energy transformation". @stap-m deleted the existing subclasses of "energy production" (which is now "energy transformation"), see issue #77. We thought it would be good to discuss the subclasses from scratch and maybe take some old definitions.

Ideas of solution

I would define the subclass processes of energy transformation as a transformation process from the primary energy source X into a secondary energy form (electricity, heat or electricity and heat).

I saw that in the OEO under continuant --> energy: there are some primary energy sources and there are also the secondary energy forms mixed togehter. I would separate them and complete the list (but that's maybe an issue by itsself).

For the subclasses of energy transformation process I would inlude:

photovoltaik plant: Transforms the primary energy source solar power into electrical energy. (+old def)
solar heat plant: Transforms the primary energy source solar power into heat. (+...)
Concentrated solar power plant: Transforms the primary energy source solar power into electricity (and heat) (+...).
wind onshore plant: ...
wind offshore plant: ...
hydro power plant:...
tidal energy power plant
biogas power plant
solid biomass power plant
geothermal power plant
coal power plant
lignite power plant
natural gas power plant
mineral oil power plant
nuclear power plant
renewable waste power plant
non renewable waste power pant

I hope I didn't forget any type of power plant...

Workflow checklist

[ ] I discussed the issue with someone else than me before working on a solution
[ ] I already read the latest version of the workflow for this repository
[ ] I added this issue to the Project 'Issues'. If suitable, I add it to further Projects.
[ ] The goal of this ontology is clear to me

I am aware that

[ ] every entry in the ontology should have a definition
[ ] classes should arise from concepts rather than from words

0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q commented 4 years ago

I would define the subclass processes of energy transformation as a transformation process from the primary energy source X into a secondary energy form (electricity, heat or electricity and heat).

There are also SE→ SE transformations (all the Power-to-X stuff). So don't limit this to PE → SE.

0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q commented 4 years ago

For the subclasses of energy transformation process I would inlude:

photovoltaik plant […]

Shouldn't processes be subclasses of energy transformation process (e.g., something like "solar radiation to electricity transformation process") which are in turn realised in all kinds of power plants (e.g., photovoltaic plant)?

l-emele commented 4 years ago

The power plants and power generating units already have quite a detailed structure as subclasses of artificial object? Why do you consider breaking this structure and changing power plants from artificial objects to processes?

Vera-IER commented 4 years ago

From an energy model point of view we need both. We need the the power plants as artificial objects to descripe for example the stock of power plants, but we need also the energy transformation as a process, where we could for example describe the efficiency of the energy transformation. The sub classes of the power plants as artificial objects should be equal to the energy transformation ones.

l-emele commented 4 years ago

So what you intend is basically what we had before with the energy technologies. We had already a lot of discussions on this, e.g. #86 #136 and #173.

Efficiency and other quantities could also be seen as properties of the power plants. We just did this with capacities #320 #391.

Vera-IER commented 4 years ago

Yes I would take the same subclasses as those which you already discussed for the power plants.

Yes true efficiency was a bad example, that could be also a property of a power plant. A better example are emission coefficients. They need to be multiplied with the amount of energy transformed/produced and can't be properties of power plants, but could be properties of energy transformation processes.

I remember that we also talked about including energy production. Maybe its also better to call that energy transformation?

l-emele commented 4 years ago

I remember that we also talked about including energy production. Maybe its also better to call that energy transformation?

We already have a class energy transformation: Energy transformation is a process in which one ore more certain types of energy as input result in certain types of energy as output. energy production could be subclass of energy transformation, i.e. that energy transformation where primary energy (carriers) come from.

(Related to #390.)

stap-m commented 4 years ago

From an energy model point of view we need both. We need the the power plants as artificial objects to descripe for example the stock of power plants, but we need also the energy transformation as a process, where we could for example describe the efficiency of the energy transformation. The sub classes of the power plants as artificial objects should be equal to the energy transformation ones.

Is it really necessary to include (however) all those types of energy transformation? Maybe it is sufficent to include all relevant parts to annotate such a transformation. For exmple the transformation inside a pv cell: input: radiation --> energy transformation --> output: electrical energy and thermal energy (losses)

Vera-IER commented 4 years ago

Sounds like a good solution. I don't know - is it possible to use the ontology like a puzzle? Can I describe a specific transformation process by combining several parts of the ontology and combine it with an attribute and a value?

@p-kuckertz do you know?

p-kuckertz commented 4 years ago

No, I'm of no use here, sorry. I do not know about this way of utilizing an ontology and I'm no domain expert...

stap-m commented 4 years ago

Is it really necessary to include (however) all those types of energy transformation? Maybe it is sufficent to include all relevant parts to annotate such a transformation. For exmple the transformation inside a pv cell: input: radiation --> energy transformation --> output: electrical energy and thermal energy (losses)

and

Sounds like a good solution. I don't know - is it possible to use the ontology like a puzzle? Can I describe a specific transformation process by combining several parts of the ontology and combine it with an attribute and a value?

@jannahastings @fabianneuhaus can you help here?

jannahastings commented 4 years ago

Sounds like a good solution. I don't know - is it possible to use the ontology like a puzzle? Can I describe a specific transformation process by combining several parts of the ontology and combine it with an attribute and a value?

Sure, this could just be captured in an anonymous class expression in an axiom rather than a named class, we already use such expressions in some cases.

Reading the thread above, I would just like to underscore that power plants and energy transformations need to be separate entities. Of course, the power plant facilitates the energy transformation.

Vera-IER commented 4 years ago

I don't understand it yet. The term "energy transformation" already exists. Can we somehow relate this energy transformation process to all existing power plants in the OEO? If yes, I think we don't need the subclasses of energy transformation?

stap-m commented 4 years ago

Of the existing relations, participates in fits quite well: a relation between a continuant and a process, in which the continuant is somehow involved in the process E.g.: powerplant participates in some energy transformation All subclasses of power plant should inherit this property.

l-emele commented 4 years ago

Not only the power plants, but all energy converting devices.

stap-m commented 4 years ago

On the other hand: it is the energy converting device that does the energy transformation process. We should better add it here. Or both.

Vera-IER commented 4 years ago

I like the idea to add this relation to all power plants and energy converting devices. Is the discussion now closed?

stap-m commented 4 years ago

I just took a look into the hierarchy: for consisenty reasons we should either add the relation only to energy converting device or additionally to energy generating unit and power plant (which are related via has part) .

Before we implement, we should wait for #77 and decide: do we relate to energy transformation or to energy generation (or how it will be called)?

l-emele commented 4 years ago

Why has part? A process is not an material object.

stap-m commented 4 years ago

Why has part? A process is not an material object.

power plant --> has part --> power generating unit --> has part --> energy converting device

l-emele commented 4 years ago

That is clear, but I think has part is not the appropriate relation between energy transformation and energy converting device. Or do I understand something wrong?

stap-m commented 4 years ago

Yes, the suggested relation between energy converting device and energy transformation is participates in :)

l-emele commented 4 years ago

Oh, sorry. participates in is fine.

k-knosala commented 4 years ago

From an energy model point of view we need both. We need the the power plants as artificial objects to descripe for example the stock of power plants, but we need also the energy transformation as a process, where we could for example describe the efficiency of the energy transformation. The sub classes of the power plants as artificial objects should be equal to the energy transformation ones.

Is it really necessary to include (however) all those types of energy transformation? Maybe it is sufficent to include all relevant parts to annotate such a transformation. For exmple the transformation inside a pv cell: input: radiation --> energy transformation --> output: electrical energy and thermal energy (losses)

For futher specification of the energy converting device I think this would be a good way do descripe all energy transfomation processes. As mentioned before in this issue and in #434 it allows to assign attributes to the transformation process (e.g. efficiency, power of transformation etc.) that are describes as process attribute.

0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q commented 4 years ago

From an energy model point of view we need both. We need the the power plants as artificial objects to descripe for example the stock of power plants, but we need also the energy transformation as a process, where we could for example describe the efficiency of the energy transformation.

So if I have two different Diesel generators, one with efficiency of 0.3 and one with efficiency of 0.29, they would be energy converting devices that participate in different energy transformation processes?

k-knosala commented 4 years ago

@jannahastings wrote in #434:

There are always four separate things: (1) the entity in reality (some sort of specifically dependent continuant, perhaps, or a process) (2) a number (3) a unit. The entity then has_value some (4) quantity value entity, that has_unit the unit. (And may have a specified value with a data property). The relationship between the quantity value entity (4) and the entity in the world (1) is is_about.

Following this chain, the energy converting device participates in energy transformation (process) has_some input, has_some output, has_value efficiency (quantity value entity) has_unit ratio. The realisation of the quantity value entity efficiency is then a data property.

Does the data property have to be directly connected to the independent continuant or is a modular approach as in this chain the way to go?

l-emele commented 4 years ago

From an energy model point of view we need both. We need the the power plants as artificial objects to descripe for example the stock of power plants, but we need also the energy transformation as a process, where we could for example describe the efficiency of the energy transformation.

So if I have two different Diesel generators, one with efficiency of 0.3 and one with efficiency of 0.29, they would be energy converting devices that participate in different energy transformation processes?

They would be different instances of the same classes.

0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q commented 4 years ago

So they would be different.

stap-m commented 4 years ago

Please continue the efficiency discussion in #434 to not mix up the issues.

0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q commented 4 years ago

I'm just wondering why there is a class process class separate to the object class when the process is only ever realised by the object. Claiming different generators of the same type implement different processes, because they differ in some attribute, strikes me at unconventional at a minimum. And I don't see yet what can be achieved with this construction that's not possible with attaching attributes directly to objects.

akleinau commented 4 years ago

I'm just wondering why there is a class process class separate to the object class when the process is only ever realised by the object. Claiming different generators of the same type implement different processes, because they differ in some attribute, strikes me at unconventional at a minimum. And I don't see yet what can be achieved with this construction that's not possible with attaching attributes directly to objects.

@0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q: @Vera-IER mentioned above emission coefficients as attributes of the process, not the objects

Yes true efficiency was a bad example, that could be also a property of a power plant. A better example are emission coefficients. They need to be multiplied with the amount of energy transformed/produced and can't be properties of power plants, but could be properties of energy transformation processes.

akleinau commented 4 years ago

So the idea is we don't make new classes and use if needed the anonym class: The energy transformation that has the energy converting devicexyzas participant. This class than can get attributes like emission coefficients. Which would mean we don't change the ontology and close the issue?

Vera-IER commented 4 years ago

Which would mean we don't change the ontology and close the issue?

Don't we need to relate the process energy transformation with participates in to all the power plant types in the ontology?

akleinau commented 4 years ago

oh yes right, so do that but don't add new classes

Vera-IER commented 4 years ago

Okay I can try to implement this. To make sure, if this is the right way to do it: I include subclasses of the power plant types e.g. coal power unit. With the object restriction creator in protege I chose the restriction property participates in and the restriction filler energy transformation and the restriction type some?

akleinau commented 4 years ago

mostly yes, but the latest idea was to give that relation to the "energy converting device" class, not the power plants. You don't need to implement for every subclass, it will get inherited from the "energy converting superclass" if that one gets the relation.

Vera-IER commented 4 years ago

If I relate it to energy converting device, it is related to heaters, generators, etc. These energy converting devices are not related to energy carriers, so it would not be possible to add attributes for example emission coefficients. It's only possible to add such attributes if I would relate it to power generating unit. So shall I implement it for power generating unit then?

Vera-IER commented 4 years ago

or for both energy converting device and power generating untit?

akleinau commented 4 years ago

I just took a look into the hierarchy: for consisenty reasons we should either add the relation only to energy converting device or additionally to energy generating unit and power plant (which are related via has part) .

Before we implement, we should wait for #77 and decide: do we relate to energy transformation or to energy generation (or how it will be called)?

@stap-m can you explain your thoughts behind energy converting device? Otherwise I'd go with power generating unit as @Vera-IER proposed

stap-m commented 4 years ago

If I relate it to energy converting device, it is related to heaters, generators, etc. These energy converting devices are not related to energy carriers, so it would not be possible to add attributes for example emission coefficients. It's only possible to add such attributes if I would relate it to power generating unit. So shall I implement it for power generating unit then?

Sorry, I don't get the connection to energy carrier and emission coefficient here. Could you give an example? I mean, " heater participates in energy transformation " should be a correct relation.

1) We have the existing connection: power generating unit has part some generator 2) We implement: energy converting device participates in energy transformation which hands it down to generator 3) Does power generating unit "inherit" the " participates in energy transformation " from its part generator @akleinau ? If not, we'd need to implement the relation also for power generating unit.

Vera-IER commented 4 years ago

Yes its definitly a correct correlation, when energy converting devices are related to energy transformation.

From my model perspective I need to give attributes like efficiencies or emission coefficients to types of power generating units. A coal power plant has a higer emission coefficient than a gas power plant for example. For example if a have X TWh power generated by a coal power plant, the model calculates the emissions by multiplying the emission coefficient with the generated power. Its related to the combustion process of a power plant (which is different for the different energy carriers), not to the generator.

akleinau commented 4 years ago

Does power generating unit "inherit" the " participates in energy transformation " from its part generator @akleinau ? If not, we'd need to implement the relation also for power generating unit.

no, it doesn't inherit it

jannahastings commented 4 years ago

We should still think about which is the more primary of these and first capture that (I assume the generator?).

The other then could be, in principle, derived, and although as @akleinau says it will not be inherited automatically, we should still think about whether we need both and if so whether there is a regular relationship between them that can be automatically computed (even if perhaps not within OWL) and if that would help?

stap-m commented 4 years ago

@Vera-IER I completely agree, that for modelling the overall efficiency or emmisions of the power plant is relevant, and not of a single generator. But these are different relations. I don't see a problem to implement such relations for power plant, regardless of this issue.

Vera-IER commented 4 years ago

Did I understand it correctly that we can implement both relations?: energy converting device participates in energy transformation power generating unit participates in energy transformation

akleinau commented 4 years ago

this issue has 45 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?

Vera-IER commented 3 years ago

In the OEO Dev Meeting we agreed to include those relations: energy converting device participates in energy transformation power generating unit participates in energy transformation power plant participates in energy transformation

I will implement it, but I don't know where to put the term tracker item in that case. Because they exist already for energy converting device and so on. Should I include a second term tracker item for theses terms or leave them out?

l-emele commented 3 years ago

Should I include a second term tracker item for theses terms or leave them out?

Please include in the term tracker items the issue #372 and the PR #646 additionally to the issues and PRs already included.

OpenEnergyPlatform / ontology