You have experienced a slow internet. Your music starts chopping. The video stops streaming. You curse the hour glass and spinning donuts. You have never experienced a substantial dip in power speed. Yes; power has a speed, and it is 60 Hertz (Hz), or 60 cycles a second. The reason I can argue it is a speed is because it is governed by spinning generation, namely steam turbines, from which we still get over half our electricity.
For-Instance Bogusness #1: Frequency Modulation
Your computer doesn’t melt down; servers don’t melt down; switches, hubs, routers – all that internet stuff will never see damage as a result of data demand exceeding supply. Guess what for electricity – equipment is going to start hurting bad with just a few percent of fluctuation from 60 Hz. Grids must be synchronized to precisely the same 60 Hz.
But even more phenomenally challenging is it isn’t just any old 60 Hertz. The first image is a metaphor for electricity delivered by the “internet of things”. These people may all be walking at roughly 1.7 Hz, or 100 steps per minute. They represent your solar panels, little generators, and other distributed energy resources (DERs). Chaos.
Instead, the grid must be in sync like Kim Jong-un’s army – stay in line or you’ll spend a month on bread and water in the gulag.
Can the crowd of people walking the street be tamed to walk in sync? Sure. This would be all the frequency regulation and paralleling switchgear and devices that are expensive and absolutely critical.
For example, modern farms require power all the time. If power is lost for more than a few minutes, livestock will start dying en masse due to suffocation – a loss of ventilation.
If I have a few hundred kW of backup generators, like all farmers do these days, and I lose grid power, I break the connection with the grid, start my generator, and turn my loads over to the generator. I don’t want or need expensive paralleling equipment.
A minute or two with no power is of no consequence. When grid power comes back, I break the connection with the generator, cutting all power to my loads again, and then flip the switch back to the grid.
Because the generator and grid are not synchronized – marching in step. What happens otherwise? Boom! Fireball. It is essentially taking voltages that could be vastly out of sync and slamming them together like two trains on the same track going in opposite directions or vastly different speeds in the same direction.
For-Instance Bogusness #2: Distributed Resources
In the 1980s, we had computer labs with rows and rows of IBM desktop computers sold by Charlie Chaplin, or rows and rows of Apples, Apple IIs, Apple IIes, or Macintoshes. They all stood alone with no connection to anything except a wall socket. Actually, prior to that, since processing power was so expensive, we had mainframes in client/server format. This is what I had to suffer through to fake and hobble my way through FORTRAN class. The psychological scars were so bad I was terrified of the Charlie Chaplin PC.
After the desktop revolution, we started networking computers because wouldn’t it be cool to share files without running floppy disks around? Then we started getting larger and larger servers for more and more data.
Now where are things headed? To the cloud. Companies are ditching servers and instead centralizing it all in somebody else’s massive data center(s). Box, Dropbox, Amazon Cloud, and Google Drive are just a few storage alternatives. Applications run in the cloud too, of course.
What does this all resemble? The conventional power grid. Exactly. It is centralized for redundancy of equipment, location, and power supply, so there is no interruption, very much like the obligation to serve under which utilities operate. It is also done for economies of scale so customers can focus on their business and not dinking around with equipment.
It makes no sense in any economic model or trend to start scattering little dinky servers all over the place and putting out bids for 20 GB of photo storage on the server in Jeff Ihnen’s basement. Why would anyone do that? Why or how is the electrical grid different in any way?
The way I see it, people can have their DERs, which are just like this Dell E7250 laptop into which I’m pounding this blog. If the power goes out and I lose internet in my house, I limp by with what I have on this machine.
The cell tower has DER to maintain power and customer service during a power outage. I connect with that DER-powered data resource with my phone and tether to my laptop. I’m back on the internet like the farmer’s DER. If my outage is longer than my device batteries can handle, I have a small gasoline-powered DER in my garage, and I can keep going. Have I done this? Arrr, you better believe it!
This is far, far different than running my own power plant with extra power to sell to the grid and having my own oversized computer capacity to lease to others. I could literally list 50 reasons why the business case for the DER energy internet of things makes no sense at all. Using DER for backup and as a demand response resource is worth considering but is expensive to put in place and deploy.
Join the discussion 3 Comments
Your piece from your series on the “Bogus Internet of Things”, specifically “For-Instance Bogusness #2 Distributed Resources” is well written, and very clearly communicates engineering insight across the disciplines to those of us who are involved like you in this energy efficiency effort but are not engineers. I really liked your examples of the crowd on the sidewalk vs. the North Korean military to illustrate frequency. This is like Feynman’s writing, freighting hard truth. Necessary, practical and, even brilliant. If engineering texts were written this way it would probably cure the drop out rate at engineering schools.
However, suppose what is happening is climate change, say rapid climate change instead of business as usual. Then, every city will need many robust micro-grids (packed with various DERs and stuffed DSM and DR) to insure that essential services (police department, fire, medical, cell phones, water system, central city administration) can operate when the grid goes down. Sandy and Katrina are initial lessons in what will happen to linear grids. The vision is independent micro-grids that can maintain service and trade with each other – the system is envisioned as an ecology made of many redundant loops rather than s a linear system. The reality is that, so far, it is as you describe. Climate adaptation micro-grids with DERs, or just DERs with integrated DSM without the micro-grid are weak. Cities of a certain size require that these things work. It is not a choice. The Center for Advanced Engineering Studies in Boise, NREL, a number of university based and other engineers and the American Society of Adaptation Planners (ASAP) are working towards solutions. I remember reading that Edison was initially big on DC. Since the urban micro-grids envisioned and in the initial pilots are small, should they be designed internally to be DC? Would that help with the frequency problem? I think most off-grid cabins are DC and internally computers are DC, but I might be wrong. What do you think?
Gil – you outstripped my knowledge of alternating and direct current options… but I did investigate for answers. We have a grid that is almost entirely AC (there is some high voltage DC transmission but very little). The reason is it is easy and cheap to step voltages up and down with AC transformers. High voltage is needed to keep conductor sizes down for long distance i-squared-r losses (resistance).
The thing is, as you would guess, everything is built for AC so using DC would seem to just require more doodads to convert it back to AC for home use. What I also learned is that DC circuit breakers are 100x as expensive as AC breakers. http://www.theenergycollective.com/rogerrethinker/204396/ac-versus-dc-powerlines
So although a lot of loads us DC power, delivering it inexpensively to the customer and then safely (with over-current protection) within the customer home/facility is likely to keep us on AC delivery for a long long time. It would be more disruptive than switching to metric! I remember that push from the late 1970s!