System load will change in the next decades, just as a consequence of the
energy transition.
The illustrations of the phases do not reflect these changes by somehow adjusting the shape of the LDC.
There are three reasons for this rude simplification:
- The first reason is a didactic one:
explaining diverse interactions, it helps to reduce complexity. If everything is changing at the same time,
what are the relationships you can observe?
- Any visual indication of expected changes would be highly speculative. An important part of the energy transition is an increase of efficiency.
This should reduce the annual load. Simultaneously, coupling of economic sectors introduces new loads.
The timelines of vRES growth, efficiency increase and sector coupling may vary a lot per country and, hence,
there is no robust relationship to be shown in the illustrations.
- Many supporters of power system transformation argue that digitalisation and smart grids in combination with flexible loads
(as well as storage) will help to match the demand with the actual available vRES generation.
This would allow to mitigate some of the challenges related phases 3 and 4.
However, flexibility and the capability to communicate / control are two very different things.
Control is a precondition to effectively deploy existing flexibility. But the nature of load (or generation)
determines the accessible level of flexibility.
That's why I am reluctant suggesting beneficial potentials whithout providing a supporting analysis.
Flexibility perspectives
Deploying flexibility potentials - regardless of load, generation or storage - it is essential to be explicit
what perspective you have in mind:
- system: balancing load and demand at whatever level;
- network: managing load flows avoiding congestion; or
- market facilitation: economically optimising portfolios and maximising profits.
Attempts to combine these perspectives regularly reveal conflicts rather than synergies.
The illustration below is inspired by the methodology of the
'Flexibility Tracker' (see slide 8 of the presentation).
It helps to understand potentials and structure of flexibility.
According to the
'Flexibility Tracker',
flexibility providers are generation, loads and storage.
However, these three groups are not completely flexible. In fact, they are quite diverse.
Generation
- vRES generation is inflexible (otherwise there was no challenge).
Wind and sun do not respond to demand or market signals.
In case of excess generation, some curtailment may help to match load,
but even this option is restricted by its economic implications.
- The flexibility of dRES generation varies with technology and resource.
Run-of-river hydro is quite inflexible. Biomass or hydro dams offer some more flexibility.
However, they are subject to restrictions as well.
Think of irrigation needs / water management, environmental regulation
or the fact that storage capacity is not indefinite.
- Flexibly dispatchable generation from other sources, often, is associated with carbon emissions.
In terms of capacity (power), gas turbines are highly beneficial for power systems with high vRES shares.
In terms of energy, though, usage should be limited as much as possible.
(Nuclear power plants may be dispatchable, but flexibility is limited.
Rather than enabling vRES integration, their dispatch easily is running into conflicts with vRES.)
Loads
Much of the total load is inflexible because of its purpose.
Digitalisation or smartness will not change this. Streetlights are needed during darkness and only during darkness.
The schedule of trains preferably does not depend on prices at power exchanges.
Of course, you can decide to postpone the start of your washing machine or dishwasher or to iron your shirts later (if you iron them at all).
Looking at your immediate living environment, you may realise that it is challenging to name many more of those examples.
Assessing flexibility of loads, it is important to be specific about the related time frames.
Some applications, after appropriate preparation, allow shifting demand for a couple of minutes (electric arc furnaces), some for hours (cooling) but very few for days or even months.
Nevertheless, these long timeframes are essential too when trying to match supply and demand.
Another aspect which deserves attention: once suspended, loads may ask for some extra power after being reconnected.
If supply is still short, this may jeopardise the benefit of flexibility.
Storage
Because of the limitations of generation and load, scenarios with high vRES penetration always assume storage.
Storage provides flexibility in time. Periods covered range from minutes to seasons.
This is determined by technology, capacity and application.
Storage can be implemented as electricity storage like batteries.
However, very often, it is attractive to deploy storage capacity in other forms of energy.
Closely related to generation, this can be water reservoirs like hydro dams or pumped storage plants.
Closer to loads, an elegant approach is using the inertia of the applications' final energy use.
Examples are heat e.g. in case of electric arc furnaces or cooling appliances or buffer storages in drinking water supply.
(In fact, these latter cases can be equally qualified as load flexibility. This is just a matter of definition.)
Dedicated electricity storage in the networks may also provide ancillary services.
In most countries far from the equator, seasonal storage is an inevitable precondition for realistically matching supply and demand.
Electrochemical batteries are not an obvious choice for this purpose, even if they get extremely cheap.
One single charge-discharge cycle per year does not justify putting huge amounts of materials in related equipment.
Export and import
In case of excess generation or a power deficit, export and import, respectively, may help to maintain the system balance.
Let's interpret exports and import wider than just transmitting power via interconnectors to other regions.
The transformation process implies closer links of the power system with other economic sectors,
like mobility or heating (power to heat) or cooling in the built environment.
Excess generation may be converted into chemical energy carriers (power to fuel).
These synthetic fuels may never be transformed back into electricity.
Nevertheless - the fact that they can be stored easily adds flexibility to the system.
On the other hand, some (zero-carbon) energy carriers need to be transformed back to power to cover deficits.
This is the import in the graph. (Of course, in general, imports will not go directly to the load
but will enter the system at the generation side.)
Exchanging power with other economic sectors can be considered as new load and generation,
as storage or as export / import - this is just a matter of definition.
Important is the extra flexibility offered by these concepts resulting in potential support for system balance.
Of course, synthetic fuels can also be physically exported, imported and traded between countries
like coal, oil or natural gas nowadays. This is an efficient option for reducing the required volume of seasonal storage.
The
'Flexibility Tracker'
characterises networks and power markets as
flexibility enablers.
It is important to understand, that proper regulation is a precondition for unchaining existing flexibility potentials,
but more networks or markets will not create them.
Again: offering money for switching streetlights on during daytime in case of excess PV generation
does not make this a sense-making application.
Networks and interconnection
While storage provides flexibility in time, networks and interconnection increase the spatial diversity and,
hence, increase the deployable flexibility in space. Interconnection offers synergies like enabling vRES integration,
increasing economic efficiency of supply and strengthening security and reliability of supply.
When planning networks for integrating vRES, there is one aspect which has to be understood:
some limited curtailment is an intrinsic part of all power systems with high shares of vRES. - Why that?
The peak output from PV plants or windfarms occurs only a few hours per year and, hence,
peak output (in terms of power) is associated with a limited share of the annual yield (in terms of energy).
It is economically efficient to chose the rating of the network assets lower
than the nameplate capacity of the connected vRES projects and accept a very limited yield loss.
What are the consequences? For safe network operation, some monitoring and control infrastructure
has to be implemented enabling the network operator to temporarily reduce vRES plant output during peak hours.
(Alternatively, the output of individual projects can be capped statically at values lower than nameplate capacity.)
From the perspective of vRES development it is crucial that the investors' risks associated
with curtailment are manageable.
Even if the total expected amount of yield loss will be low, for bankability and financing of projects
it is important that regulation covers and reliably limits the financial risk to predictable levels.
Power markets
Economic incentives stimulate efficient use of existing flexibility.
However, liberalised power markets do not create flexibility, neither are they a precondition for successful power system transformation.
Important is adequate allocation of costs and benefits - otherwise nobody will be willing to provide the needed investments.
Investments at the right spots: this is how markets - on the long run - can increase flexibility of power systems.
This website focuses on technical challenges.
Elaborating on appropriate design of economic frameworks for vRES development
(and power markets as one possible implementation) is another ambitious web-project. One day you may find a link here...
