If you drive through the bustling oil patch of the Eagle Ford Shale near San Antonio, it won’t take long to find the surreal sight of flares burning natural gas like perpetual bonfires.
Natural gas is cheap. Pipelines are expensive. So instead of collecting the fossil fuel, many oil and gas operators build tall, metallic spires called flare stacks to burn the gas and release it into the Texas sky.
For years, no one could say with any certainty how much natural gas was going to waste. Everyone knew flaring in shale country was a problem. But officials at the Railroad Commission of Texas, the state agency that oversees the oil and gas industry, had never released figures showing how much was being burned in the Eagle Ford.
Instead, the agency released only statewide figures showing the overall volume of flaring was low compared to overall production — about one percent.
Whenever a government agency touts rosy statistics, there’s probably a database behind those numbers. And if you obtain that raw data, you might be able to figure out what’s really going on.
Today’s Express-News story about flares burning 20 billion cubic feet of natural gas so far in 2014 is a good reminder of the value of public databases — and why journalists need to get their hands on them to analyze the records for themselves.
There’s no question analyzing data can be a lot of work. We filed an open records request with the Railroad Commission for a copy of the flaring data in the spring of 2013. It’s a huge database of monthly reports showing how much oil and gas is produced in Texas and where those hydrocarbons go. Flaring and venting are one of the “disposition” categories in the data.
I drove to the agency’s Austin headquarters with a flash drive that could handle the enormous database. It was a beast — more than 25 gigabytes of 85 million records. All that summer we used software to convert the Railroad Commission’s archaic data to CSV files, a format we could use in the newsroom. After that, it took weeks to crunch the numbers and uncover the hidden pitfalls.
Why go through the hassle? Why should frazzled journalists take the time to learn how to analyze data? Don’t we have enough to do?
The answers is, journalists need to know a lot of skills — how to interview people, how to write clearly, how to find information. Analyzing public data should be a part of that skill set. It opens doors to stories that couldn’t otherwise be told. This is what journalism is all about.
When we were finished reviewing the flaring data, our analysis showed that the volume of flared gas in Texas had increased by 400 percent since 2009. And most of that gas came from the Eagle Ford Shale near San Antonio. This chart essentially told the story of flaring in the shale that no one had figured out — not even state officials:
Quantifying the volume of flared gas opened up new questions and possibilities. When Projects Editor David Sheppard asked how much air pollution was created by all this flaring, we found out there was a way to calculate an estimate. We obtained emails from the state’s environmental agency, the Texas Commission on Environmental Quality, that showed how to estimate levels of air pollution created by gas flares. Those formulas were based on the volume of flared gas – which we had. So we plugged those numbers into Excel spreadsheets to come up with the amounts of sulfur, volatile organic compounds and other pollutants that came from flaring in the region.
In August, the Express-News published the results of our investigation, Up in Flames. The total volume of wasted gas in the shale from 2009 to 2012 was almost 39 billion cubic feet — enough to meet the annual heating and cooking needs for all 335,700 residential customers who relied on gas last year in CPS Energy’s service area, which includes San Antonio.
The new numbers for 2013 and 2014 show that flares burned and wasted even more of the fossil fuel. In the first seven months of 2014, more than 20 billion cubic feet of gas went up in smoke — enough to fuel CPS Energy’s 800 megawatt Rio Nogales power plant during the same time frame.