Energy Industry Structure and its Changes
Its time to look at another industry, energy generation and more generally generation of electricity.
Energy Industry Structure
https://energia.fi/en/energy_sector_in_finland/energy_networks/electricity_networks
The traditional energy networks are highly hierarchical and centralised. Electricity is generated in large power stations typically using hydro, nuclear or fossil fuels like coal or natural gas. Energy is distributed nationwide at high voltage in main grid at 400 kV, 220 kV and 110 kV) managed by the transmission system operator (Fingrid in Finland). From the national grid electricity gets distributed to users via distribution networks. The distribution networks are at 20, 10, 1 or 0.4 kilovolts. The lowest voltages of up to 1 kilovolt are called low voltage, the higher voltages are medium voltage (1–70 kilovolts) or high voltage (110–400 kilovolts). Primary transformers have for the longest time been WAN connected into the grid operators management system (called Scada systems). Different systems and teams are responsible for monitoring grid condition and different set of tools and people responding to faults and fixing issues.
Homes get their energy from the low voltage grid, some large industrial, commerce or agriculture users have direct connection to mid voltage or even directly to the national grid.
In Finland the industry is separated into electricity generation, national grid and regional distribution service operators (DSOs). The split is due to an EU regulation. In Finland there are 76 DSOs, some other EU countries have decided to have one national grid operator and one DSO. Latter one is an innovative way to implement regulations without implementing them.
Towards Decentralisation
This is now changing from a very hierarchically organised grid into a more distributed system where many of small energy providers get connected and sell their generation capacity through the common distribution networks. At the same time the use of electricity is increasing as it gradually replaces fossil fuels. This means that the energy system in future needs to be still reliable to ensure functioning society but its complexity is increasing all the time.
In order to combat global warming, governments are promoting use of renewable energy through subsidies – mostly for solar and wind. Other sources being developed are wave and almost none is put into development of geothermal. Solar, wind and wave have very different generation profile than traditional power stations. The renewables change their output power very fast and have big seasonal differences. There is also less economics of scale in solar. Generating power at your own rooftop costs not that much more than doing it in dedicated solar power installation.
Weather impacts heavily what is the best electricity generation option. In Finland there is very little renewables generation during winter months as an example. Long rainy seasons in warmer climates pose similar problems where sunlight can be limited to a few hours daily.
This leads to a new structure where there are large number of distributed nodes in the grid that from time to time generate a lot of power and on other times none. These nodes can be both consumers and products – called prosumers – and they swing from production to consumption fast based on changing weather conditions.
This also means that the needs for communications traffic increase fast. First to smaller transformer fields at DSOs and finally to park level transformers. Today connections to primary transformer fields are from a single WAN provider. In future all transformer fields will have remote connections, local networks and local edge processing capacity. Satellite connections for backup etc. Local processing capacity will be partially done to do video analytics to identify strange travellers in and around fields, but also for running new applications (analytics etc.) to increase stability of the networks.
Grid Control Through Price Signals
Prosumers make independent decisions when to sell and buy just based on their need and electricity generation at the moment (or they have software automating those decisions). The end result is that there will be at periodically more energy generation than consumption - for example on very sunny days. At the same time some hours will not have enough to cover the need. And these can be the same day.
Earlier energy generation was adjusted based on demand but as the energy generation becomes dependent on weather, in future the demand must adapt to the generation also.
The demand is expected to be controlled primarily with price signal. Spot price goes up or down based on demand causing users to decrease or increase usage on short cycle and electricity producers to increase or decrease production on a longer cycle.
This changes how the grid is managed.
The grid is one connected entity where electricity flows in alternate currency and the frequency and phase need to be kept within tolerances. For example, the frequency in Finland between 49,0Hz and 50,1 Hz. In the grid electricity does not flow along a single shortest path between producer and consumer but flows along parallel lines as well. Each line has maximum capacity and the amount of power over each line must be below its capacity. Exceeding capacity causes lines to heat, sag and may cause breaks or phase and voltage fluctuations.
To combat this the national grid operator needs to ensure a balance between production and consumption by buying adjustment capacity from markets and paying for it (ability to quickly increase production or reduce usage). During the coldest winter days when there is no wind and sun does not produce anything, iron mills and other factories can earn more by shutting down their production than from actual operations. A perfect time to do short maintenance if possible.
Basically, you have one huge machine with severe constraints (capacity of lines and transformers cannot be exceeded), and its frequency and phase must be kept in tight tolerances. Into it come a large set of independent, small electricity generation where each prosumer decides independently whether to use capacity, load their batteries or sell to the grid. This is a setup where in future tens or hundreds of thousands of ‘managers’ turn the knobs and push the levels of a shared machine without any co-ordination other than price signal. And the grid operator fixes everything after the fact. Since the grid has worked well in centralised configuration, this is expected to work beautifully in decentralised net as well.
The fluctuation of energy generation leads to large swings in energy price. Some hours the energy price may be negative – users are paid to use electricity – and at other times the prices can jump tenfold. If left to market forces, the large energy spikes would cause commercial providers to start building solutions to capture the value (store energy when it is cheap or paid for and sell when prices raise). Political dynamics in many countries are likely to interfere to curb the price peaks or create compensations if price goes very high for certain user groups etc. This slows down or prevents commercial solutions and will lead to a more unstable energy network.
Some centralised power generation will be dismantled due to governmental subsidies for renewable energy or plain bad policies. Governments are using tax payer money for added unreliability in electricity networks.
Pricing
The nature of the new grid will also affect pricing. Today consumers are charged based on consumed electricity but in a network with fast swings in generation this becomes increasingly unsuitable. In future consumers will be charged both based on the maximum agreed or peak power used as well on the total consumption and the power-based charging component will become more important.
Demand Response Systems
There are ways to mitigate sudden drops in generation capacity such as limiting or ‘shaping’ energy use or to have energy generation capacity that can quickly kick in. An additional solution is to build connections to neighbouring countries so that extra amounts of electricity can be bought form the regional electricity markets. It levels local changes in generation capacity but does not solve seasonal changes (winter/summer).
Demand response is one good way to shape the peaks of energy use. Large energy users have made contracts with national grid operators that they can cut their energy use during peaks for a compensation.
On consumers side energy companies can offer monetary compensation to users who are willing to let the demand response provider reduce temporarily their power usage as long as it does not affect living comfort. Such uses could be for example cutting air conditioning or direct electric heating, stopping electric car charging for short periods etc.
A demand response provider aggregates such consumers together and can offer sizeable ‘NegaWatts’ to the grid operator. When there is need to curb power consumption, the demand response provider will rotate short usage reductions between groups of consumers to achieve the needed reduction. Such virtual power station company may also sell electricity from its users’ home storages – batteries at home or in electric cars.
Load shaping can happen both on national or regional level. On local level the capacity of an old feeding transformer may cause problems that require shaping the demand even when there is no shortage on national level.
If local manufacturing with 3D printers and other scaled down production facilities becomes common, it becomes possible to create the virtual factory. Virtual factory is any production that utilises the grid only when price is low enough. It may have some own generation capacity or simply work on times when price is right. Or it may let customers decide. Some customers who can wait, may order products to be manufactured only when price is negative or at zero.
Energy Storages
The energy imbalances can be mitigated with energy storage as well. Many solutions exist and are being developed. Different technologies suit different needs.
Superconductor and high-speed rotating masses (flywheels) can respond very fast and can prevent short time outages. They act like UPS (uninterrupted power supply) systems but for the grid. These still need maturation. For example, spinning large steel disks at high speed in strong magnetic fields start to lose their rigidity and managing the material becomes very difficult.
Large scale grid batteries can respond also fast and store more energy. Large scale batteries can in addition be used to run various quality improving functions such as voltage support, provide reactive power to grid etc. There are two main technologies here – Lithium-ion and flow batteries. The grid batteries have quite different needs than batteries in cars or phones. Grid batteries need to store large volumes of energy at low cost whereas in car batteries for example the need is to store very fast and have a light weight. Grid batteries do not have these limitations, they can tolerate slow charge and large weight. This leads to differences in technology choices.
Reserve gas or diesel generator-based electricity generation can respond within a few minutes and are good solution when the need is longer. For example, to balance between summer and winter but also for short term outages of tens of minutes to hours. This can be natural gas or synthetic fuels generated via so called power-to-x solutions that we’ll return later. Synthetic fuels would be methanol, methane and ammonium. Ammonium has the good quality that it contains no carbon so using it in a modified combustion engine generates no CO2 releases.
Another option is to break water in electrolysis to hydrogen and oxygen and use the hydrogen later in fuel cells to turn back to electricity. This is an area that has recently gained a lot of popularity. The costs for changing to a hydrogen based system are high as hydrogen gas cannot be pumped using existing natural gas infrastructure pipes. As the smallest molecule in universe, hydrogen gas (H2) is the smallest molecule in universe can can escape through normal steel structures in between the atoms. Hydrogen needs its separate national pipeline systems.
Pumping water to upper water reservoirs offers largest reserve capacity that can be turned back to electricity with conventional hydropower station when needed. This requires that the terrain has big differences in elevation and is geographically restricted. It has been used among others in Alp region in Europe. Some old and deep mines can find new life as power storages.
Other potential solutions include aggregating consumer electric storages or electric cars with their batteries that can be used to temporarily release electricity back to the grid.
The table below summarises the main characteristics of the alternatives