Are we living in fairy land?
- Energy blind.
- Smil.
- Hanson, evolutionary unstable strategy.
Intro to Modelling Course
- Backcasting is a form of validation. Take past data and run the model with that data and compare to the historical outputs.
- Given a set of assumptions, this is what the power prices will be. So it’s not so much about predicting the future generally.
- Why do we need a power market model? Power markets are complicated. If you’re investing in this market you want quantitative analysis for your decisions.
- inputs:
- Demand assumptions are probably some of the biggest assumptions. Developed by the research teams.
- Commodities assumptions. Come from global commodity markets. The commodities team generate these from other models.
- Weather patterns. Demand profile is heavily influenced by weather patterns. Take a weather pattern from a historical year and use it as an assumption for future years.
- Policy assumptions. Varies heavily based on the region. Fairly open ended from the modelling teams.
- There is a granularity to the inputs. In a spatial, temporal and categorical sense. e.g. The weather pattern of a specific region, half-hourly and filtering on types of power plants.
- outputs:
- Prices. Wholesale prices of renewables. Capture prices?
- Capacity and generation. The model is allowed to decide whether to build or retire plants. Does this based on forecasting incentives to build in the sector. So clients are interested in what capacity is ‘still standing’.
- Capture price vs wholesale price for a technology? How much money you’ll make for your generation.
- System level. Demand, imports and emissions.
- The outputs are stored in Data Warehouse and can be accessed with power bi, sql and origin explorer.
- soul:
- Iterative loop between dispatch and scaricity. Dispatch to power markets, allow the model to make build decisions and contracts to drive investment.
- Demand + Charging + Exports = Production + Imports + loss of load (can’t meet demand, blackout, known as slack variable). This is done every hour. So it’s trying to meet demand in every time period.
- The market clears in the cheapest possible way. So look at the supply you have in a period for your cheapest energy source. Done using a ‘supply stack’.
- The wholesale price is determined by demand and cost to supply energy. Just assuming that margins aren’t being made. That theres enough competition such that running costs is your market bid.
- Uplift is applied to output. To close the gap between the real world and model output.
- Decides on building based on net present value of a plant.
- The model makes decisions based on what will minimise total system costs over the lifetime of the model.
- CapX costs are an input.
- The solve is performed over the complete time line. The model has complete knowledge.
- Sample Run: sample a set of days in a year. Do this to lock in build decisions and capacities.
- Use thes build decisions on a full year run, using all days. Just dispatching.
- Capacities are used across regions. So teams share the output capcities of model for given regions. The inputs are then fed together for other regions. The other regions you feel are important for the current region run.
19/10
I’ve been somewhat inspired by Werner Vogels’ go build motto these days. Just that enthuasiam to not be nihilistic about tech, to look at it’s positives.
I also need to change my thinking about constraints in terms of what can and can’t be done. Obviously its easier said than done but identifying the correct bottlenecks is important.
I’ve thought about the idea of the IoT future and that. Not sure what I mean.
Listened to https://www.youtube.com/watch?v=t3UwyHjz8BU&t=4385s&ab_channel=SikaStrength It’s kinda humbling to realise that I’m not that special in terms of training. In a comparative aspect, you’ve to focus on yourself and your goals. Aesthetics is not the goal. Or maybe it is, just don’t know if its a healthy one.
23/10
I don’t think I weight the impact of policy and just culture in determining peoples choices. For instance, in eating meat. The planet is a massive farm. And there is reason for this of course but there’s no inherint reason we can’t change now.
Reading: https://ourworldindata.org/land-use-diets
Wabi Sabi.
Accepting the flaws, rawness and natural beauty. The imperfections of substance and sound in their most natural raw form.
24/10
I listened to The Proof podcast w/ Hannah Ritchie this morning. She says the environmental goal of our food choices is to reduce land use for agriculture.
beef has an energy efficiency of about 2%. This means that for every 100 kilocalories you feed a cow, you only get 2 kilocalories of beef back.
Smaller animals are generally more efficient. It takes less energy to keep them alive.
Simon Hill also makes a good point. To establish the principles of the diet you’re eating first. To slowly move into it step by step, knowing the why each time.
Listened to Lyn Alden on money as well. She makes this distinction between the ledger and the unit of account of that ledger. The ledger relies on the kind of polictical dimension, social acceptance and that. The unit of account is a seperate dimension and having gold as it reduces the risk of the ‘ledger controllers’ debasing the value of that particular dimension.
There’s some good fundamental numbers and graphs in here to add to Anki.
In the hypothetical scenario in which the entire world adopted a vegan diet the researchers estimate that our total agricultural land use would shrink from 4.1 billion hectares to 1 billion hectares. A reduction of 75%. That’s equal to an area the size of North America and Brazil combined.
Fundamentally, animals are really good at converting energy dense but low protein quality cereals into high-quality protein.
If the use of cereals across the world is broken down into three categories animal feed, human consumption and industrial use. The human consumption component is less that 50% (48%) worldwide and close to 17% in Ireland.
When he was young, he was struck by watching a panther move in a cartoon. That this drawing was a cat moving, but not just any panther, a ‘stuffy, cranky’ panther. That all these attributes were conveyed in the animation. It’s fascinating to think that you could have such a deep insight at such a young age.
Spent a lot of time consuming. What he liked, particularly why he liked it or why he didn’t. Try and figure out what you would do to fix it once you’ve discovered the why.
Engagement sometimes is difficult. I’ve been looking into AWS networking today but I’m not engaging with the problem that it is trying to solve.
Agriculture has impact in 3 ways:
- Large fresh water use.
- Large amount of land use.
- Responsible for around 25% of global GHG emissions.
Need to understand the Energy Problem.
Build hardware, building a computer for your problem.
Oppurtunity cost of land use. Without cattle, the carbon sequestration of the land alone is greater than having a livestock ecosytem. And the subtract the carbon cost of protein by alternative means.
30/10
I listened to Tim Spector on Rich Roll this morning. I got to say, I’m not entirely convinced about the reductionism of nutrition. This is largely because it stresses me out a bit so I’m not saying it too generally. But I just don’t know how much stock we should really place in anything but longer term nutrional advice.
Current way to establish understanding
- Look at PNAS consensus.
- Investigate major studies.
I also listened to George Manbiot over the weekend. I agree, superficially, with his notion that culturally we inherentely want to dominate the landscape. He makes a decent case that farming is considered inherentely good, having roots in religion.
I think reductionism plays a role, subtely, in a lot of areas I study. For instance, music. You’re looking for this lineage, dividing lines between genres and periods as opposed to starting with the holistic, what you like and investigating why you like it.
I think generally, I’ve not weighted highly enough how important just optimism and believing things can change and the future can be crafted.