And yet, people in the business are claiming that natural gas wellheads should never freeze.  "Aflatoxin" at PeakOil.com said:

Natural gas wells are probably not going to freeze.

They use Dehydrators at the well heads or CDP, and just in case, there is plenty of Methyl alcohol. Once the gas makes it into the interstate pipeline system, it is bone dry, about 95% methane, 2% N2, 1.5% CO2 and a little ethane and heavies.

Water is a big problem in pipelines, the gas people worked out ways to remove all of it 50 years ago.

so your saying the freezing story was a cover story rather then admiting a sudden production shortfall?
I honestly don't know.  The Denver Post paints a much more complex picture:


Mechanical malfunctions, inaccurate weather forecasts and inadequate natural-gas supplies caused the rolling power outages that afflicted 325,000 customers of Xcel Energy on Saturday.

It marked the first time in Xcel's institutional memory that controlled electrical outages occurred in the winter.

Such disruptions typically happen on the hottest summer days when demand for power peaks.

Three outages of 30 minutes each hit customers in portions of metro Denver, Grand Junction and the central mountains. The outages lasted longer than 30 minutes for about 25,000 customers.

"The phrase 'perfect storm' is overworked, but we've never seen events like this," said Xcel spokesman Mark Stutz.

Portions of three coal-fired power plants in Colorado were out of commission Saturday morning, either for scheduled maintenance or because of mechanical breakdowns.

In addition, Xcel had inadequate supplies of natural gas on hand because initial weather forecasts had not suggested the record-breaking low of minus 13 on Saturday.

The utility had enough natural gas for all of its customers' heating needs, but it ran short of the fuel to supply gas-fired power generators that supplement the electricity produced by coal-fired power plants.

The gas shortage was exacerbated by an undetermined number of Rocky Mountain natural-gas wells whose pumping equipment froze, preventing additional deliveries.

The idea that one natural gas production equipment failure could cause Xcel to need to take down the grid would make sense ONLY if there were very little spare supply.

Also, this stuff about 30 minute blackouts isn't terribly accurate. It was more like 45 minutes to an hour. And it wasn't until later in the day that the cause, a natural gas shortage, made it into the media.

It is not gas that may have froze, but water condensation that slowly built up.  Eventually the passage for the gas narrowed.  Until that passage was cleared, getting gas to the customer became more difficult. (That is how I understood the "freezing.")

However, the story did mention a shortfall, as well as the freezing.

Someone from Denver should address this issue.

As I understand it, there should be no water in the natural gas at that point.  They have hundreds of dehydrators at each wellhead, as well as other ways to remove water and anything else that might freeze.
There usually is only 1 to 3 dehydrators to a field, which could have many wells, often up to 50 miles away from the last well in the gathering network. I worked a gathering system in south Texas for a field of 860 wells. We had 3 dehydrators at the 3 outlets from the gathering system network all about 50-60 miles distant from eachother. The wells had enough pressure to flow to those outlet points by themselves, but required a number of 1500 HP compressors to kick up the pressure enough to get into the intrastate gas transmission company's lateral lines. That's where we put the compressors. Compressors and flow meters can't take liquid slugs, so all liquids had to be knocked out beforehand. After the liquids are out, you can run the gas through a scrubber to eliminate the H2S (if you have any of that nasty stuff), then dehydrate the rest of the gas stream, compress it and lastly, meter the flow to the transmission company, so it was much more convenient to minimize the installation all of that equipment to the fewest number of places as possible.
The gathering field lines freeze between the wellheads and the first glycol dehydrator, which in many cases can be up to 50 miles away from a well.  In the gathering lines there can be a mixture of oil, water, salts, condensates, paraffins, waxes, NG, CO2, N2 maybe H2S.  Hydrates are crystals where water is embedded in gas molecules.  Any significant quantity of hydrates can plug the gathering lines.  Hydrates can form at temperatures above 32ºF.  If temperatures drop below 32ºF, then ice plugs can also form.  A lot of wellhead treater installations have heaters to keep the flow into the gathering lines above the temperatures where hydrates form.  Normally the well heaters are off unless hydrate formation temperatures are expected, in which case an operator will go out to the field and light up the heaters.  Even if the heaters are lit, they can still go off if hydrates form upstream, as this can shut down their fuel supply which also comes from the well.  In my experience it sometimes not possible to reach the wellheads in time to light the heaters and keep them from freezing, because of downed trees, and snow/ice on the roads and access trails.  It is also not uncommon for somebody to  get sick on those cold nights either.  Of course extremely cold weather usually happens on the weekends when there is only one person on call.
Someone who claims to have worked directly for this particular company says they have hundreds of dehydrators per wellhead.  He didn't think it was likely that anything would freeze, even with equipment failure.
If he worked for the local distribution company, or the intrastate or interstate company, he is probably correct. In those systems the gas would not be subject to freezing. If he/she says they h/s worked at the company operating the gas fields, do a detailed check his/her credentials. The company I worked for was one of the largest independent drilling and gathering operators in the US at that time (we had more gas wells and gas flow volumes than Exxon produced by themselves) at the time. The only other thing that could possibly have happened (within my reasoning) is that the Denver local distribution company (LDC) had thier peak gas diverted by their intra/interstate supplier to other customers that were experiencing greater hardships or had contracted for more firm transport capacity than Denver. Normally, as was suggested by other posts, various customers will request a firm volume that the intra/interstate transmission company must guarantee to transport and another peak volume that the intra/interstate transmission company will transport only if he can reasonably do so. They will follow that SOP unless an emergency is declared, in which case they will revert to making special volume assignments according to priority level of the services that are affected. Could have been one of those things. I've said all about I can say without having been sitting in the control room myself, so ... I'll definitely leave any further details for the LDC to explain away.
He says he worked for CIG, before it was acquired by  El Paso.

He also said all of CIG's wells are in steep decline, except the Jonah field in Wyoming.  And their plants are not in great shape, either.  

He's right then.  I never heard of Xcel. I left the USA in 1990 and there's been a lot of takeovers since then, but I do remember Colorado Interstate Gas and most of their immediate supply wells were in the Tx-Ok panhandles and SW Kansas.  Those were the old gathering systems in which I was trying to turn one around to supply a proposed peak generator in Larime Co.  They were some producing wells still left around that particular area, but they were on their last breath 16 years ago and were running at about 30-40% N2, so even then they wern't what you would call star performers.
But he's still wrong about having hundreds of glycol treaters.  There's only a couple of those per field.
Xcel used to be called Public Service Company.
The only XL that I heard about before was Microsoft's.  I'm surprized they don't sue Excel for some name infrigement violations.-)