What’s the Difference Between a Grid and a ‘Net?

Are electricity grids like the Internet? Not enough

Historically, a great deal of growth has come from advances in one area being picked up and expanded on in another. Over the past 30 years, the Internet has made huge progress, and it makes sense to look at whether some of those advances could be copied on electricity grids.

Stepping back, there are some huge similarities between a grid and the internet and some major differences. What are they?

How electricity grids work

In some ways, electricity grids remain close to what we all learned in school. Simplified, they are made up of just a handful of components:

• There are generators that produce electrical energy (usually by spinning magnets inside coils of wire, sometimes using photovoltaic cells).
• There are consumer systems or devices in houses and businesses that want to convert that electrical energy into some other form. E.g. motors, light bulbs, or heaters.
• There are networks of wires of varying forms that physically connect them.

There is also storage capacity. For example:

• Hydroelectric pumped storage.
• Some grid-scale batteries.

Outside of the grid there is also storage in consumer batteries like those in mobile phones.

However, electricity storage is still small potatoes compared to the amount flowing on the grid.

Recently, things have got more complicated because households and others can be generators and consumers and provide storage due to all those rooftop solar panels and house batteries. However, things haven’t fundamentally changed that much. It’s mostly still about wires of one form or another connecting producers to consumers. The wires have become a lot more swanky in design and how they’re used but they’re essentially still wires.

That somewhat undersells the storage aspect of the grid itself, which I’ll talk about in a moment, but first let’s take a look at the Internet.

What is the Internet?

The Internet doesn’t deliver electrical energy. It delivers data. Nevertheless, there are plenty of similarities between it and an electrical grid.

The Internet is often described as, “wires, storage, and compute”. Using the same model, electricity grids could be described as “wires with a tiny bit of storage,” and one of their major differences is that storage and compute play a much greater role in the modern Internet than in any public grid. Although, it’s worth remembering compute and storage were far less of a factor in the early internet.

One reason for the current situation is that data storage is much cheaper than energy storage. However, it’s also because clever compute has made data storage more effective, and it is now a sophisticated contributor to the overall effectiveness and stability of the Internet. Basically, the internet has learned how to use compute and storage to manage and shape consumer demand. That played a key role in the internet’s survival during the 2020 pandemic.

Routing

The Internet and electricity grids are both complex delivery networks with payloads that have to be routed between creators and consumers.

On the internet, that payload is a discrete data packet and the routing is done actively using protocols such as BGP (the one that is always famously cutting off access to huge chunks of the net). The power on electricity grids, however, is different.

Electricity isn’t routed in the same way as data packets because electricity isn’t discrete, it is continuous. In most cases, it is already present in the wires outside your house and, unless something goes wrong, the only thing that ever changes about it is the frequency or voltage in those wires.

The exception is when you open a new path to the electricity, in which case it will flow along that pristine new “wire” at the speed of light. That’s what happens when you turn on a new device like a light bulb.

As grid and smart meter expert Mat Roderick describes it, “Electricity is pervasive, and it’s like a moth seeking flames, electricity seeks new wires and new things that consume it.”

Resistance isn’t futile

That doesn’t mean we have no ability to steer the power on the grid. Electricity follows the path of least resistance and the resistance of particular transmission lines can be artificially increased or decreased to route power towards areas where there is demand for it or away from areas where there isn’t.

What is Power?

When new power is generated, say by a solar panel, and flows into the local grid the amount of energy in that grid increases. Electricity is an electromagnetic wave, so that means at least one of two potential two things happens: the frequency of the electricity will go up or the voltage (the wave’s amplitude) will.

Grids operate within a range of frequencies and voltages. For example, the UK grid operates at a frequency of 50Hz, but supports +/-1% fluctuations around that as business as usual. Devices attached to the grid are tolerant of the frequencies being higher or lower than the target 50Hz (newer ones tend to be more tolerant than older ones)

This means that by increasing the frequency the grid itself can provide a form of temporary energy storage. This storage is not permanent — if a substation or generator failed there’d instantly be no power in the wires — but there is the ability to retain energy for short periods.

This means that when high demand is predictable, such as the classic example of a lot of kettles getting turned on at the end of the World Cup, the grid can be fed with extra energy in anticipation, resulting in higher frequencies. As this power is drawn out to heat the kettles, the frequencies drop back again to lower levels.

In fact, frequency monitoring is something modern grids use widely to check and moderate the power available in particular areas.

Too much or not enough?

So, couldn’t the grid be used as a giant battery by letting the frequency rise and rise!? What could go wrong?

Even if all devices could operate at higher frequencies, not all of the grid could cope. Some lines would burn out, which would result in more power in the remaining lines, and that could cause a set of cascading failures. The irony is, the real danger to a grid is not too little power, which might result in low frequencies and some devices not working, but too much, which might result in stuff blowing. Sometimes important stuff.

“As electricity gets stretched, it thins out, reducing frequency and volts maybe to the point where older devices don’t work with it.

If it has nowhere to go it gets fat and angry, increasing frequency and voltage until things start to blow (and that destruction will cause blackouts)” — Mat Roderick

Get the balance right

Grid balancing is the process of making sure the right amount of energy exists in the right parts of the grid to match the demand in those areas and the local capacity to handle the power. Grid balancing in specific locations takes the form of:

• bringing sources of power on or offline,
• increasing or reducing demand,
• routing power towards or away.

The goal is to maintain frequencies and voltages within their target ranges so no devices fail and nothing gets overloaded and blows. If the entire grid was a perfect superconductor, it would balance itself and no wires would ever get overloaded. Unfortunately, it isn’t. Many parts of the grid need to be protected from too much power overheating them.

The problem is, none of these grid balancing actions are fast-acting. Although the directed changing of demand (Demand Side Response or DSR) has the potential to be quick it is, so far, only available in a limited form.

Unfortunately, this means the grid cannot react rapidly to changing conditions even if those new conditions are immediately visible via rising or falling frequencies. The grid is forced to predict what it can, and act on those predictions, with hopefully enough slack in the system to handle unanticipated events.

Working on projections of the future is just about good enough for now. However, the situation is about to become more unpredictable. Fossil fuel based power can be forecast well in advance, but weather-dependent renewables are less predictable. That will be tricky to manage unless the grid can control demand in a more responsive fashion. That is the aim of DSR.

Demand side response

The good thing about batteries is they are a source of high demand. House batteries and car batteries are often connected to the grid, which means they could potentially be a source of grid balance. Switching them off is a way to reduce demand or they could draw excess energy out of a grid at short notice — if they could be communicated with quickly enough.

There are other potential sources of demand: dishwashers, heaters, washing machines, dryers. These draw less power than batteries but there are a heck of a lot of them.

On a Tightrope

As time goes on, the energy on the grid is going to become harder and harder to balance using the current mechanisms. We need new ones that are faster and fed by better data. Data needs to be more timely, local, and granular and it needs to be in the hands of devices that can act on it intelligently to stabilise the grid, either by drawing more power or less of it.

The internet has taught us effective intelligence requires data, compute, storage, and the ability to take rapid action. Devices on the edge of the grid such as smart car chargers provide compute, storage (potentially virtually in the form of delayed demand), local data, and agency. What they don’t have at the moment is grid data.

That’s why we need to get it to them.

Photo by Armand Khoury on Unsplash