6 ms·

Programmers occasionally have envy that civil engineering is "real engineering", e.g. civil engineers know how to build a bridge and deliver it on time. Wherea

by chubot 14y ago

Programmers occasionally have envy that civil engineering is "real engineering", e.g. civil engineers know how to build a bridge and deliver it on time. Whereas software engineers will come up with 1000 ways to build the bridge, 10 of which work, and take 10x longer to build it. This article seems like a good counterexample. Like most failures, this is a systems failure. Someone engineered the hell out of that cable. But they neglected to realize that it would be drawn through a steel pipe. "So, what went wrong? Varying load conditions in the three legs of the 3-phase circuit caused tremendously strong and dynamic magnetic field changes. The electromagnetic forces between the three conductors and the steel pipe (gack!) cause the conduit to wiggle around inside the pipe." The other lesson is that you shouldn't hyper-optimize locally, and instead design for maintainability. I'm not sure what they were thinking when laying a cable that requires experts to be flown in from the east coast to repair. The simplest mechanisms are the ones that are the most easily repaired. The entire article goes on in great length about how awe-inspiring the engineering of this cable is, whereas I just see a bunch of complexity that probably didn't need to be there. The 118 layers of paper tape, and the oil pumps, and so forth. Surely there was a simpler solution. I'm sure after this failure and the ones on the east coast, they are not still laying cables like this.

mikeash 14y ago
I don't know about this automatic rejection of complexity. It could very well be that, even accounting for the tough learning curve and the cost of repairs, it's still the cheapest way to transmit that much power in that context. Sometimes massive infrastructure requires massive repairs, but that doesn't automatically make it not cost-effective.
pixl97 14y ago
At the time the cable was installed (72ish?) 230kV was a very high voltage cable. The type of cable that now commonly carries that voltage underground is HV XLPE http://www.fujikura.com.sg/hv_xlpe.htm http://www.fujikura.com.sg/hv_xlpe.htm At the the maximum voltage over XLPE was 110kV. Sticking 2 of these cables side by side would have other interesting effects. http://www.emfs.info/Sources+of+EMFs/Underground/multiple.htm http://www.emfs.info/Sources+of+EMFs/Underground/multiple.ht... It appears oil and paper systems are still used, but I don't think they use the high pressure oil system like the one in the paper. http://www.hitachi-cable.com/about/publish/review/__icsFiles/afieldfile/2005/11/28/1_review02.pdf http://www.hitachi-cable.com/about/publish/review/__icsFiles...
elteto 14y ago
There is no "simple mechanism" to transport 600-800 amps across 10 miles. That amount of current alone is just mind blowing.
- rdl 14y ago
  I think this would depend on voltage. Transporting 800 amps 10 miles is done all the time -- a normal house has 100-200A service. The issue is that this is 230 KV, vs. 230V.
  - elteto 14y ago
    I didn't know this, I can see how a house appliance might consume several dozens of amps but 100 to 200 amps? Wow, that is a lot. I don't have much knowledge about power transmission (which is one of the most fascinating fields in electrical engineering, in my opinion) so I'll take your word for it.
    - rdl 14y ago
      You can have a small crappy house/apartment with 40-60A service, for sure. But electric appliances (hot water heaters, clothes dryers, ovens/ranges, etc.) can be 50A. A single excellent commercial electric wok is about 50-100A. Each individual circuit is about 15-20A. The really awesome thing now seems to be HVDC transmission -- eliminates capacitance problems and problems with multiple grids.
      - paulgerhardt 14y ago
        > The really awesome thing now seems to be HVDC transmission -- eliminates capacitance problems and problems with multiple grids. This. Imagine a global grid connecting alternative energy sources for each optimal locale: Geothermal in Iceland, Solar in North Africa, and of course Thorium in Norway. Peter Thiel seems to think this scenario is improbable[1] but I don't think it takes into account ABB's recent breakthrough with HVDC switching[2]. [1] http://blakemasters.tumblr.com/post/23787022006/peter-thiels-cs183-startup-class-14-notes-essay http://blakemasters.tumblr.com/post/23787022006/peter-thiels... [2] http://new.abb.com/about/hvdc-grid http://new.abb.com/about/hvdc-grid
pavel_lishin 14y ago
> e.g. civil engineers know how to build a bridge and deliver it on time. Tell that to the Texas Department of Transportation, where any given highway project will typically run over time and budget.
- angersock 14y ago
  To be fair, all those Aggies have to end up somewhere. :)
- 205guy 14y ago
  I believe that is a feature and not a bug. At least for the contractors and their friends in governmnet.
purephase 14y ago
An always interesting facet to these types of stories is the ambivalence I feel for the people involved. While I would love to have a problem to tackle that is large enough, similarly unique enough to solve that it would make my career, I would have a problem large enough, similarly unique enough to break my career. As another commenter mentioned, the engineering involved in the story is likely not required very many places on earth. To be in a situation, developing a solution to a massive problem with little to no precedence is an exhilarating and terrifying position to be in. It may sound overly-engineered to you but, at the time, it may have been the most cost effective engineering solution to a problem rarely encountered.
sbierwagen 14y ago
The entire article goes on in great length about how awe-inspiring the engineering of this cable is, whereas I just see a bunch of complexity that probably didn't need to be there. The 118 layers of paper tape, and the oil pumps, and so forth. Surely there was a simpler solution. http://www.howtospotapsychopath.com/2010/03/10/today-on-surf-celebrity-science-class/ http://www.howtospotapsychopath.com/2010/03/10/today-on-surf... Take care that you don't come down with "Engineers' Disease", though - the tendency for people with a high level of technical knowledge to decide that their knowledge must be applicable to specialised fields that they don't actually know a lot about. You are a programmer, not an electrical engineer. Wild speculation regarding a field you know little about is not attractive. Allow me to reverse your statement. "I don't see why programming for multicore processors has to be so complex. Why not just run four copies of the program on all four cores?" "I don't see why every web site makes you create a new account. Why not log into everything with my facebook account?" "I don't see why the legal system has to be so complex. Why not just have the judge decide everything, and get rid of the lawyers?"
- pestaa 14y ago
  Calling out parent's incompetency is fair I guess, although we still don't know why it was necessary to engineer a cable so much only to find out it has basic flaws, hence the assumption it is over-engineered in certain aspects and under-engineered in others.
- jrockway 14y ago
  "I don't see why programming for multicore processors has to be so complex. Why not just run four copies of the program on all four cores?" This is actually a good idea.
  - whatshisface 14y ago
    That only works if your program is a one-way pipe for data, like a graphics card. Most programs need to communicate within themselves.
    - 14y ago
mhb 14y ago
Chesterton's thoughts on those ready to reform that about which they know little: http://stuartbuck.blogspot.com/2004/05/chesterton_14.html http://stuartbuck.blogspot.com/2004/05/chesterton_14.html
greenyoda 14y ago
"So, what went wrong? Varying load conditions in the three legs of the 3-phase circuit caused tremendously strong and dynamic magnetic field changes. The electromagnetic forces between the three conductors and the steel pipe (gack!) cause the conduit to wiggle around inside the pipe." This is sort of like the electrical engineering equivalent of the Tacoma Narrows Bridge collapse: https://en.wikipedia.org/wiki/Tacoma_Narrows_Bridge_(1940)#Collapse https://en.wikipedia.org/wiki/Tacoma_Narrows_Bridge_(1940)#C...
greenyoda 14y ago
"I'm not sure what they were thinking when laying a cable that requires experts to be flown in from the east coast to repair. The simplest mechanisms are the ones that are the most easily repaired." In this case, the simplest mechanism would probably have been to build above-ground high tension power lines on massive steel towers. This is the most common mode of high voltage electrical transmission, and the cables are easy to access and repair. However, this cable was probably built when the area was already densely populated, and property owners would not have accepted having such an eyesore going through their back yards. So they had to bury it underground, which required heroic engineering to insure that it would only have to be maintained very rarely (designed for an MTBF of 60 years). There are analogous problems in computer hardware. For example, if you want a computer to control a cardiac pacemaker, you can't just buy a generic server and throw a gigabyte-sized Linux distribution on it. It has to fit in a tiny space, run for years on a small battery and be 100% reliable. Very different from creating a web application, but the extra complexity and expense are necessary due to the problem that's being solved.
- giardini 14y ago
  "In this case, the simplest mechanism would probably have been to build above-ground high tension power lines on massive steel towers. This is the most common mode of high voltage electrical transmission, and the cables are easy to access and repair. " astonishing that they don't do this. Their first attempt failed; the second likely will also. And both at tremendous expense when a tried-and-true methodology exists.
- ZeroGravitas 14y ago
  "easy to access and repair" in this context might involve ckimbing out of a helicopter onto live half million volt lines in a space age chainmail suit: https://www.youtube.com/watch?v=9tzga6qAaBA&feature=youtube_gdata_player https://www.youtube.com/watch?v=9tzga6qAaBA&feature=yout...
imgabe 14y ago
> Someone engineered the hell out of that cable. But they neglected to realize that it would be drawn through a steel pipe. This is about the most arrogant, ignorant thing I've read on HN. No. I promise you the electrical engineer who designed that feeder did not "neglect to realize" that a steel pipe is conductive and if the feeder came into contact with it, it would short. There were no doubt a number of considerations and trade-offs involved in the choice of material.

by chubot 14y ago

mikeash 14y ago

I don't know about this automatic rejection of complexity. It could very well be that, even accounting for the tough learning curve and the cost of repairs, it's still the cheapest way to transmit that much power in that context. Sometimes massive infrastructure requires massive repairs, but that doesn't automatically make it not cost-effective.

pixl97 14y ago

At the time the cable was installed (72ish?) 230kV was a very high voltage cable. The type of cable that now commonly carries that voltage underground is HV XLPE http://www.fujikura.com.sg/hv_xlpe.htm http://www.fujikura.com.sg/hv_xlpe.htm At the the maximum voltage over XLPE was 110kV. Sticking 2 of these cables side by side would have other interesting effects. http://www.emfs.info/Sources+of+EMFs/Underground/multiple.htm http://www.emfs.info/Sources+of+EMFs/Underground/multiple.ht... It appears oil and paper systems are still used, but I don't think they use the high pressure oil system like the one in the paper. http://www.hitachi-cable.com/about/publish/review/__icsFiles/afieldfile/2005/11/28/1_review02.pdf http://www.hitachi-cable.com/about/publish/review/__icsFiles...

elteto 14y ago

There is no "simple mechanism" to transport 600-800 amps across 10 miles. That amount of current alone is just mind blowing.

rdl 14y ago
I think this would depend on voltage. Transporting 800 amps 10 miles is done all the time -- a normal house has 100-200A service. The issue is that this is 230 KV, vs. 230V.
- elteto 14y ago
  I didn't know this, I can see how a house appliance might consume several dozens of amps but 100 to 200 amps? Wow, that is a lot. I don't have much knowledge about power transmission (which is one of the most fascinating fields in electrical engineering, in my opinion) so I'll take your word for it.
  - rdl 14y ago
    You can have a small crappy house/apartment with 40-60A service, for sure. But electric appliances (hot water heaters, clothes dryers, ovens/ranges, etc.) can be 50A. A single excellent commercial electric wok is about 50-100A. Each individual circuit is about 15-20A. The really awesome thing now seems to be HVDC transmission -- eliminates capacitance problems and problems with multiple grids.
    - paulgerhardt 14y ago
      > The really awesome thing now seems to be HVDC transmission -- eliminates capacitance problems and problems with multiple grids. This. Imagine a global grid connecting alternative energy sources for each optimal locale: Geothermal in Iceland, Solar in North Africa, and of course Thorium in Norway. Peter Thiel seems to think this scenario is improbable[1] but I don't think it takes into account ABB's recent breakthrough with HVDC switching[2]. [1] http://blakemasters.tumblr.com/post/23787022006/peter-thiels-cs183-startup-class-14-notes-essay http://blakemasters.tumblr.com/post/23787022006/peter-thiels... [2] http://new.abb.com/about/hvdc-grid http://new.abb.com/about/hvdc-grid

pavel_lishin 14y ago

> e.g. civil engineers know how to build a bridge and deliver it on time. Tell that to the Texas Department of Transportation, where any given highway project will typically run over time and budget.

angersock 14y ago
To be fair, all those Aggies have to end up somewhere. :)
205guy 14y ago
I believe that is a feature and not a bug. At least for the contractors and their friends in governmnet.

purephase 14y ago

An always interesting facet to these types of stories is the ambivalence I feel for the people involved. While I would love to have a problem to tackle that is large enough, similarly unique enough to solve that it would make my career, I would have a problem large enough, similarly unique enough to break my career. As another commenter mentioned, the engineering involved in the story is likely not required very many places on earth. To be in a situation, developing a solution to a massive problem with little to no precedence is an exhilarating and terrifying position to be in. It may sound overly-engineered to you but, at the time, it may have been the most cost effective engineering solution to a problem rarely encountered.

sbierwagen 14y ago

The entire article goes on in great length about how awe-inspiring the engineering of this cable is, whereas I just see a bunch of complexity that probably didn't need to be there. The 118 layers of paper tape, and the oil pumps, and so forth. Surely there was a simpler solution. http://www.howtospotapsychopath.com/2010/03/10/today-on-surf-celebrity-science-class/ http://www.howtospotapsychopath.com/2010/03/10/today-on-surf... Take care that you don't come down with "Engineers' Disease", though - the tendency for people with a high level of technical knowledge to decide that their knowledge must be applicable to specialised fields that they don't actually know a lot about. You are a programmer, not an electrical engineer. Wild speculation regarding a field you know little about is not attractive. Allow me to reverse your statement. "I don't see why programming for multicore processors has to be so complex. Why not just run four copies of the program on all four cores?" "I don't see why every web site makes you create a new account. Why not log into everything with my facebook account?" "I don't see why the legal system has to be so complex. Why not just have the judge decide everything, and get rid of the lawyers?"

pestaa 14y ago
Calling out parent's incompetency is fair I guess, although we still don't know why it was necessary to engineer a cable so much only to find out it has basic flaws, hence the assumption it is over-engineered in certain aspects and under-engineered in others.
jrockway 14y ago
"I don't see why programming for multicore processors has to be so complex. Why not just run four copies of the program on all four cores?" This is actually a good idea.
- whatshisface 14y ago
  That only works if your program is a one-way pipe for data, like a graphics card. Most programs need to communicate within themselves.
  - 14y ago

mhb 14y ago

Chesterton's thoughts on those ready to reform that about which they know little: http://stuartbuck.blogspot.com/2004/05/chesterton_14.html http://stuartbuck.blogspot.com/2004/05/chesterton_14.html

greenyoda 14y ago

"So, what went wrong? Varying load conditions in the three legs of the 3-phase circuit caused tremendously strong and dynamic magnetic field changes. The electromagnetic forces between the three conductors and the steel pipe (gack!) cause the conduit to wiggle around inside the pipe." This is sort of like the electrical engineering equivalent of the Tacoma Narrows Bridge collapse: https://en.wikipedia.org/wiki/Tacoma_Narrows_Bridge_(1940)#Collapse https://en.wikipedia.org/wiki/Tacoma_Narrows_Bridge_(1940)#C...

greenyoda 14y ago

"I'm not sure what they were thinking when laying a cable that requires experts to be flown in from the east coast to repair. The simplest mechanisms are the ones that are the most easily repaired." In this case, the simplest mechanism would probably have been to build above-ground high tension power lines on massive steel towers. This is the most common mode of high voltage electrical transmission, and the cables are easy to access and repair. However, this cable was probably built when the area was already densely populated, and property owners would not have accepted having such an eyesore going through their back yards. So they had to bury it underground, which required heroic engineering to insure that it would only have to be maintained very rarely (designed for an MTBF of 60 years). There are analogous problems in computer hardware. For example, if you want a computer to control a cardiac pacemaker, you can't just buy a generic server and throw a gigabyte-sized Linux distribution on it. It has to fit in a tiny space, run for years on a small battery and be 100% reliable. Very different from creating a web application, but the extra complexity and expense are necessary due to the problem that's being solved.

giardini 14y ago
"In this case, the simplest mechanism would probably have been to build above-ground high tension power lines on massive steel towers. This is the most common mode of high voltage electrical transmission, and the cables are easy to access and repair. " astonishing that they don't do this. Their first attempt failed; the second likely will also. And both at tremendous expense when a tried-and-true methodology exists.
ZeroGravitas 14y ago
"easy to access and repair" in this context might involve ckimbing out of a helicopter onto live half million volt lines in a space age chainmail suit: https://www.youtube.com/watch?v=9tzga6qAaBA&feature=youtube_gdata_player https://www.youtube.com/watch?v=9tzga6qAaBA&feature=yout...

imgabe 14y ago

> Someone engineered the hell out of that cable. But they neglected to realize that it would be drawn through a steel pipe. This is about the most arrogant, ignorant thing I've read on HN. No. I promise you the electrical engineer who designed that feeder did not "neglect to realize" that a steel pipe is conductive and if the feeder came into contact with it, it would short. There were no doubt a number of considerations and trade-offs involved in the choice of material.