In the region where I live there's a state standard that says that mains voltage deviation can be within 5 percent continuously and within 10 percent for short periods of time, so if the mains voltage is within those ranges - it's just okay. The nominal voltage is 220 volts, so it can be in 209..231 volts range continuously and in 198..242 volts range for short periods of time.

Now I understand that sometimes there're undersized wires and huge losses and bad wire joints and this can cause undervoltage at the consumer site.

What would cause overvoltage? I mean there're carefully designed generators somewhere that rotate at carefully monitored "right" speeds and produce carefully precomputed voltage. Then there're transformers that again have the right number of winds in each windings and so convert the right voltage into the other right voltage. So I don't see how voltage would suddenly get higher than designed. Yet there's even a state standard that allows for rather huge deviations.

What exactly causes overvoltages in power grid?


7 Answers 7


Why is the mains voltage generally above the nominal value? I am not talking about power spikes, which leave the margins. We are talking about standard operations. By design, the power is set closer to the top margin than to the middle. These are the reasons:

Standard power generators all run with a certain rotational speed which is synchronized with the grid frequency. The rotational frequency of the generator also depends on how many poles it is equipped with, all 4-pole generators in 50Hz grids run with 1500/min, for instance.

Grid frequency is just about the only persistently constant value you can expect from the grid.

At the fixed speed, the power output of a generator is regulated by the excitation of the field coils and the mechanical input at the turbine or engine. Both values must be regulated in unison. If you increase the excitation without increasing mechanical input, the machine will slow down, and come out of sync, which must be prevented.

Some kinds of power plants run asynchronous (flywheel, solar, wind mostly) which means their power output has to be electronically regulated to fit it onto the grid.

For several reasons the power suppliers will regulate towards the upper end.

First, they can react more quickly to reduce power output: Divert some steam, reduce excitation, done. To react upwards, they must first make more steam, which takes time. So it is safer to be on the top limit.

Secondly, the same power can be more efficiently be transported when the Voltage is higher. Losses almost exclusively come from current, higher voltage means less current, so less loss, bigger percentage of voltage arrives at the customer, and only power that arrives will be paid.

Lastly, a part of the used power is pure electrical resistance, which consumes more power with higher voltage, leading to higher consumption and higher sales. I suppose this is not a big deal.

Now the power suppliers know very well how much power will be consumed on average. They know how much more will be needed on special days like thanksgiving (every stove is in action that day), or on superbowl day. They will plan ahead for quite a while.

The quality of the grid lines is taken into consideration here: If they know the voltage drop within a neighborhood rather high, the supply to that neighborhood will be set up so the planned voltage arrives at the customers, if possible. Transformators between the high/medium/low voltage networks can be regulated to some degree. (see ULTC at http://en.wikipedia.org/wiki/Tap_%28transformer%29)

Therefore voltage drops and also phase shifts are the bane of the suppliers: These two factors lead to bigger losses in the lines, which they have to pay for themselves.


You are correct that the grid is finely tuned, however it is not so static as that would lead you to believe. The entire grid is an immense machine that is quite unstable. Constant monitoring and re-adjustment is required so that the system maintains stable operations.

While you are correct that a generator generates a stable voltage (for the most part) the load on the grid changes each second. Systems monitoring these changes cannot always react instantaneously, especially when big moving object such as generators are involved.

Lets start at your home. The transformer supplying your area has three phases. The city/town planner would have devised the houses in your neighbourhood into (almost) equal amount on each phase. Now if the loads differ, it will cause minor shifts in the voltages in each phase as the phases become unbalanced. This is usually minor but can easily cause the minor fluctuations you see. If you can graph measurements over time, it should be interesting how the fluctuations look during peak times (mornings and evenings).

There are many other ways in which the grid is dynamic: Transmission lines heat up and cool down changing their resistances, solar activity induces currents in transmission lines, entire towns get knocked off the grid due to an accident. My personal favourite instability is generator phase. Generators need to be kept in phase and at frequency, however when the load on them (the grid) changes, it causes the generator to slightly speed up or slow down. This is compensated by with reaction wheels that release and absorb energy from the generator.

All of the above changes the load on the grid and therefore you will see voltage fluctuations.


As others have said, the basic problem is that demand can change quickly, but the large machines that generate electricity and the power input to them can't be changed that quickly.

Here in the US, the standard is that everything is re-evaluated every 4 seconds. The control center for each region monitors the currents thru the various transmission lines, voltages in various places, and power being dumped onto the grid by each of the large producers.

The characterisitics of each producer is known, and every 4 seconds they are told whether to regulate their power output up or down. Nuclear plants are the slowest to react, and are usually kept at "base" load. Then there are "peaking" plants that can react much faster, but also make electricity more expensively. Peaking plants are often turboshaft engines running a generator. These are usually kept off except during high demand. Hydro plants have their own sets of characteristics. They can react fairly quickly, on the order of a minute or a few minutes, to large demand changes. 4 seconds was chosen in part because at the time nothing could respond that fast. The central controller that sends out the signals every 4 seconds also applies a fairness algorithm. For example, if there are several peaking plants in the area, it tries to utilize them about equally. Managing the grid is a complex problem, and there is a lot of money to be wasted by getting it wrong.

There is a local company, Beacon Power, that makes flywheel storage systems for the grid. These are large flywheels in evacuated chambers riding on magnetic bearings. Each flywheel can store about 100 kWh of electricity. This is purely storage, not generation, but the advantage is the storing and retrieving of power is handled electronically, and can therefore react very quickly. It is possible to make a business case for a installation of these flywheels soley for the short term peaking, both absorbing and producing, they provide. Some newer power generation facilities will incorporate such short term storage locally. That lets the overall installation look like a well behaved, flexible, and fast-reacting power station, even if the ultimate power source is hydro, coal, or oil.

There is another interesting plant near hear called the Northfield Mountain Reservoir. It is a much larger energy storage station that works on potential energy of water. During light loads when the slow reacting power stations are producing more than necessary, water is pumped from the Connecticut River uphill to the Northfield Mountain Reservoir. During high demand, water flows downhill back to the river and produces power. The station has 4 reversible generators, each rated for 270 MW, so the whole station can deliver over 1 GW peak power for a while.


More or less what they said in most cases. Plus:

It takes finite time to change the power output if very large machines. Hydro turbine valves need to be opened or shut affecting tons of flowing water.Steam turbines with coal fed boilers must deal with the energy in the furnace if load drops - or have extra fuel added if load suddenly jumps.

Lighting strikes / a car hits a pole / a house fire or a broken line short a feeder. Breakers open. The fault may not propagate up the chain, or may somewhat. Load drops suddenly. Rotating machine controllers call for energy input shutdown. Water feed to turbine drops, coal feed to fire lowers ... . Voltage rises rapidly and then settles back towards steady state.

NZ and France are 12-11 just before half time in the Rugby world cup final.The ball arcs towards the goal posts - and bounces off. No penalty awarded. The ref blows his whistle and the two teams jog off the field. 1,300,000 NZers stop watching TV. 22% go the the lavatory. The water supply pump station will not notice the surge for some minutes. 127,000 electric jugs are turned on for a quick cup of coffee. More. Power load increases drastically. Voltage drops. More water is dialled up. more coal, more ... . The two teams run onto the field, kettles click off. Lights are turned off. Toilets are vacated. ... Load drops. Coal is still being added, so far ... . Voltage rises ... .

  • 2
    \$\begingroup\$ The phantom downvoter strikes again. Why not share your pearls of wisdom re what SHOULD have been in the answer - or what doesn't belong, or is wrong. \$\endgroup\$
    – Russell McMahon
    Commented Nov 25, 2011 at 14:29
  • \$\begingroup\$ This is a good answer. What you described happens every day and is the main reason for changing load, human activity. You have my upvote sir ;) \$\endgroup\$
    – Konsalik
    Commented Nov 25, 2011 at 17:43

All these generators generate exact voltages that they are built for .. it is what happens along the way.. from the generator to your plug for most part.

  • In South Africa during electric storms, lighting will strike near or direct to a high voltage line causing massacre at the step down stations- there is protection for this(and it tries to react straight away) but many times people from towns that are close will fill electrical repairs shops the next day because their tv blew up. These spikes ripple down the network which are allowed to due to tolerance levels of 10%. ( I speak from experience and not making things up here)

  • In other parts of the world, caused by hurricanes, earthquakes.

  • In other circumstances it could be caused by a tree falling onto high power lines

  • Sudden change in atmospheric properties.

  • Power grid redirection (maintanace calls)

  • But also it can caused within the home it self by devices generating feed back.

Over the years and with allot of new wiring laws introduced these dips/peaks have been mostly removed. But the tolerance is still there and most end user devices tolerate this deviation because the current is further refined using transformers in the device.

  • \$\begingroup\$ How would a blown up TV be repaired I wonder?... \$\endgroup\$
    – sharptooth
    Commented Nov 24, 2011 at 12:32
  • 1
    \$\begingroup\$ By blown up that is what customers call it- It was usually the diode on the power supply circuit that fried and gave a nasty smell and smoke. replaced for 50 bucks a pop and sorted.Jobs a good'n! \$\endgroup\$
    – Piotr Kula
    Commented Nov 24, 2011 at 12:50

As all others have said, the grid is a constant changing thing. I've seen some documentaries about local power companies here in The Netherlands. The most common thing you hear is they have a 'typical' peak periods on which they have to produce electricity. Usually power stations prepare for these moments; is there enough capacity to keep up with the rising demand?

It even goes so far that some energy companies watch the weather radar for (especially unexpected) rain, showers etc. What it so happens is that rain cools a lot of buildings down which in turn requires energy to keep them up to temperature. The typical (i.e. average) response is that people are going to use more electricity and power to keep everything warm. To counter this, the power station prepares for more capacity when it seems like it is going to rain because they know they will have to deliver more energy as usual.

All these effects are controlled by computers. A lot of statictics and 'typical expected' curves under certain circumstances are likely modelled to keep the grid from somewhat stable. Actually, only a few operators are on the power plants themselves. There may be 1-2 technicians in the small power plant itself and 1-2 operators in the office.

Coming back to your question: it's very hard to keep the grid stable. Because of the load that can change faster than the machines , a lot of the regulating is done on 'expected patterns'. Adding wind turbines to the grid makes regulating somewhat more difficult, as they can produce a few extra MW's when the wind is blowing strong, and a few minutes later it can be gone when it stops.


Main reason for over voltages are

  1. Lightning
  2. Switching surges
  3. Insulation failure
  4. resonance

Loads are resistive , inductive and capacitive in nature. in this inductive and capacitive are loads reactive in nature while resistive load are called Real (power). In a normal running power system real power and reactive power should be in equilibrium, ie (roughly) real power generated = real power consumed (load + Losses) else speed of generator & frequency will increase or decrease. Similarly reactive power generated = reactive power consumed else voltage will increase and decrease. Normally generators are equipped to adjust the real and reactive power as per load requirement by monitoring Voltage and frequency. Activities like lightning switching will cause sudden variation which results in over voltages. Inductance opposes change in current. when switching happens inductance tries to maintain the current by increasing the voltage which result in overvoltage. for further reference.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.