# How does the electrical grid cope with small fluctuations like flipping a light switch?

95 views

The operation of an electrical grid is a balancing act – you have to produce *exactly* the same amount of energy as is needed, otherwise bad things happen. I don’t understand how does this rule apply on anything other than the largest scale of things.

I understand that *in general* you can predict a higher load on the grid during Monday evening when everyone is at home, and plan your energy production accordingly. But a power station can’t predict smaller load fluctuations like if I decide to turn on or off my TV at this very moment.

So, how does the electrical grid cope with unpredictable load that differs from the planned & expected one?

In: Engineering

Teams of people at power providers assess, generally every hour and every day, the load being requested by consumers and supply electricity immediately to maintain the requested consumption.

Basically it comes down to size. Sometimes when you turn on a large appliance the lights will dim a little. That’s because you’re drawing too much power from a small section of the grid.

Now multiply that one house by thousands and homes and buisnesses the grid as a whole won’t be affected by turning on the garbage disposal in your house.

The other factor is that devices have a range that they work at. We say it’s 120v and 60hz but the voltage and frequency isn’t always the same. Enough lightswitches may cause the grid to dip a little, but it’s not enough to notice the difference.

The vast amount of energy in the grid is barely affected by a single lightbulb, so it’s just not noticeable at that level.

You’re correct that the grid is a balancing act, but the energy is equally shared across all appliances that are plugged into it. As long as the total outgoing is less than the total generated, the appliances will function correctly. If too many appliances were switch on at the same time though, all the appliances would suddenly stop working correctly until generators could ramp up and this is called a “brown out”. All the appliances are getting power, but it’s too little power. In a “black out” the grid has gone offline entirely and no appliances are getting power at all.

>So, how does the electrical grid cope with unpredictable load that differs from the planned & expected one?

It’s not really the “grid”, it’s the generators. Generators are not just generating X units of energy and that’s all there is. Rather, it’s more like when you’re running on a treadmill and the difficulty of the treadmill is changed. If the difficulty is increased, you might slow down for a moment but you quickly adjust and speed up again by putting in more effort. Generators also adjust their “effort” according to demand. Problems only happen when the demand changes so much that the generator can’t adjust to match it.

EDIT: Just to clarify, appliances (and I mean anything that uses electricity, not just toasters) are engineered to accept a range of power, not one specific amount of power. The input can dip down a certain amount and the appliance will continue to work just fine. It’s only a problem if the input varies outside the engineered range.

Imagine it like this. The power grid is like a very large lake. Turning on a light in your house is like taking out a teaspoon of water. That small quantity is not enough to make that large of a difference. But if all the lights, all the air conditioners and all the motors come on at one time, that’s like opening a flood gate.

Detailed forecasts are used to approximate what generation is needed to accommodate load on a real-time basis. It’s not perfect but it’s ballpark close.

You don’t have to produce the exact amount as what’s consumed, most generating stations aren’t operating at capacity, if demand is close to exceeding capacity additional facilities are added to the grid.

Your fridge that draws 16A at 120V which is about .3 A at 5000V which is a common voltage for a generating station to put out. Or about 0.00025% of a typical generating stations capcity of 600MW.

The electric grid doesn’t need to respond to changes as small as you turning a single thing on or off. So many people in your area are constantly turning things on and off that these tiny fluctuations get lost in the noise. The things that the grid does need to respond to are the big trends. The UK has a really interesting example of this, so many people brew a cup of tea in an electric kettle that their grid has a demand spike as everyone gets home from work and makes a cup of tea between 5 and 6 PM. It takes a nation of people all plugging something in at the same time to cause the grid to have to respond quickly.

There is a “gross level” balancing act happening. But, you are speaking of a small instantaneous change. I think we can agree the generating facilities cannot change *power* output instantly – most being some form of mechanical thing with momentum.

However, roughly, we can think of *power* as :

*Volts x Amps = Power*

So, in a given instant, if *power* is constant (since the generating systems cannot instantly change), then the other two variables are what change. In particular, adding a load to a system draws more *Amps* from the system. If *Power* is to remain constant, then *Voltage* must decrease in the instant that more *Amps* are consumed.

Therefore, adding a load to a system creates a *voltage drop* in the system. All other active devices in the system see this voltage drop and thus receive slightly less power in that instant. The “new” active device is effectively “stealing” power from all other devices in the system in that instant. If the new load is tiny (like a 100w light bulb in a house), then the impact is effectively not noticeable. Also, the effect impacts devices closest to the load more. So, turning on a heavy appliance in a house may cause lights to dim in that house (voltage drop), but not in adjacent houses.

Ultimately, if enough loads are turned on, the loads are “felt” mechanically at the generators in the form of additional rotational resistance inside the generators. This force tries to slow the rotation of the generators, but governors in the generators are designed to keep them rotating at a constant RPM. The governors signal the power plant to ramp-up “steam” to the generator to meet the load.

There are also inductors (transformers) and capacitors in the electrical transmission and distribution systems that can “store” power that smooth-out any sort of minor disturbance.

Most of our power comes from synchronous generators spun by a turbine that’s fed hot gasses from some heat source (gas turbine from natural gas, steam turbine for coal/nuclear, gas and steam turbine for combined cyclee.t.c).

You spin a synchronous generator at the frequency of your main supply (or a fraction of and have more coils in the generator, e.g half speed but twice the number of coils). The synchronous generator contains two spinning magnetic fields – one in the rotor (the bit that spins), one in the stator (the bit that doesn’t spin). When generating power, the magnetic field in the rotor is slightly ahead of the field in the stator. This induces a current in the stator and that’s where electricity comes from. The more ahead the rotor is from the stator’s magnetic field the greater the current and the more torque it takes to spin the generator.

So, when you suddenly demand a lot of power – the magnetic field in the stator slips back a bit, this increases the current to match the demand and the torque demand by the generator goes up. This causes the generator to slow down. Fortunately the inertia of the huge generators and turbines used in power generation is so large that there’s a lot of kinetic energy stored up. This means the speed doesn’t drop much and the power station has a chance to catch up by turning up the heat. In a decent grid based power network, many generators are run synchronously across many power stations and they all share the increase in demand. In some countries (including some areas of the USA) a sudden demand change may hit a few power stations very hard and they can’t react fast enough. This can cause huge blackouts as the load starts to rapidly propagate to other power stations taking down a large chunk of the power network. It can rake a while to spin it all back up again.

In the case of you flicking a lightswitch – there will be an immeasurably small decrease generator speed. The control systems in power stations are extremely accurately controlled to ensure an exact number of generator rotations in a day – and as many people turn on lights, the controller will sense that the speed is drifting fractions of a percent and act accordingly.

Im going to handwave away the electrical part and instead look at a generator and a motor and treat the power lines as a solid shaft between them.

A generator is spinning at a certain speed. You connect a motor on your end, and it spins up to the same speed. That instant you connect, the generator slows down as your motor spins up. The generator is much much much bigger than your motor, so the difference is more or less negligible. Your motor has friction and losses, it doesn’t just spin infinitely, so it requires the generator to push to keep spinning at that speed. If the speed drops too much, the power plant outputs more power to the generator, causing it to speed back up.

Everyone is connected to the same spinning shaft, so there is a ton of inertia. The power plants can very precisely determine the speed of their generator, so they can apply more power if it is going too slow, or less power if it is spinning too fast, and all the power plants are working together to maintain this same speed.

The shaft is the power lines and the speed is the voltage/frequency of the cables. There are gear boxes/transformers that change the voltage but everyone is still locked together. The force slowing down and speeding up the shaft is current.

And that is where my analogy and knowledge start to break down. But fundamentally, the reason you can flip a switch without crashing the electrical grid is immense amounts of inertia in the electrical system compared to the tiny power draw of whatever any individual electrical device might pull.

Size:

When you start your car engine, the battery can’t feed the lamps or radio. Now, connect 1000 cars in parallel and every car can now start and stop the engine randomly and the load per battery stays more or less the same.

Imagine every electric plant is a car, and every engine starter is a factory. To achieve the 1000 cars effect, you need a big grid with 100s plants. You don’t rely onto your own region or even nation. The entire Europe is cross-connected so the entire continent averages out. It even benefit of 5 time zones, so even the day-night cycle is averaged.

Last: each plant is sized to be efficient at 75% output. If the average request goes up, you just throttle some plants to 100%, you waste some efficiency but the grid will hold.

Very last: old plants are kept in “reserve”, almost shut, but ready to go if an emergency arise.