Current featured articles for the home page

Oil and Gas Explosions Are Fairly Common

On Monday morning, a man was killed by an explosion at an oil well in Bolivar, Ohio. The man is believed to have been an employee working on the site, but his identity won’t be released until it is confirmed with dental records.

This wasn’t big news in Pittsburgh, even though Bolivar is just a two hour drive from here. But why not? Is it because the incident was across state lines, or because tragedies of this sort are actually fairly routine? The answer, I think, is “both”.

In yesterday’s Pipeline, the Post-Gazette reported on a story of President Obama talking energy policy in Cincinnati. This is hardly comparable, because the words of the President are routinely discussed in national and international media. The same is not true of accidents, even those leading to fatalities, unless the number of victims or the amount of property damage is exceptionally high.

I’m not suggesting that every incident that leads to a fatality is necessarily deserving of nationwide coverage, but in some cases, the model of regional coverage can keep people from realizing that dangerous patterns exist.

As I was trying to research the incident, I kept finding more and more of them, some of which I was already aware of, some of which I was not. Here are a few examples from the past two years:

A gas explosion occurred in Northeast Philly in Jan. 2011. A firefighter moves a hose line at the scene. (Steven M. Falk / Staff Photographer) (Joshua Mellman)

  • San Bruno, CA-September 9, 2010 A 30 inch pipeline exploded, killing eight, destroying 38 properties, and damaging many more. After checking several sources, I could not find a total number of injuries. The blast left a crater 167 feet long by 27 feet wide by 40 feet deep. PG&E blamed the 2010 blast on a strength test conducted on the pipe in 1956.  Reporters covering the story initially thought the fireball might have been due to a plane crash.
  • McKean County, PA-December 12, 2010 and February 28, 2011 In separate incidents, two houses with a few miles of each other exploded without warning. The Pennsylvania DEP suspected the methane migration was due to, abandoned wells in the area, the closest of which was drilled in 1881.
  • Philadelphia, PA-January 18, 2011 A Philadelphia Gas Works employee was killed and five others were injured in this blast. The workers were trying to repair a broken gas main when a furnace glow plug ignited vapors inside a building. (Photo right)
  • Allentown, PA-February 10, 2011 Five were killed and about a dozen more were injured in a giant blast and fire that destroyed eight properties and damaged 47 others. As of this February, investigators were not close to explaining the cause of the explosion.
  • Hanoverton, OH-February 10, 2011 On the same night as the deadly Allentown blast, there was a pipeline explosion in this Ohio town. One building was damaged, but nobody was hurt in the explosion and subsequent fire that could be seen for miles.
  • Avella, PA-March 25, 2011 Three workers were hospitalized when storage tanks exploded and caught fire when a volatile vapor was somehow ignited at this natural gas well site.
  • Glouster, OH-November 16, 2011 This pipeline explosion was so strong it was felt 12 miles away. Three houses and a barn were destroyed in the blast, and one woman was hospitalized, but there was no word of fatalities.
  • Springville, PA No injuries were reported at this compressor station blast in northeastern Pennsylvania, but it blew a hole in the roof of the facility and was felt a half mile away.
  • Norphlet, AR-May 21, 2012 Three workers were killed in this blast near El Dorado, Arkansas, which according to the US Chemical Safety Board (CSB), was set off while doing “hot” work such as welding or cutting in an area with hazardous vapors.

    CSB Chairman Rafael Moure-Eraso said, “This unfortunate tragedy in Arkansas involving the deaths of three workers is the kind of hot work accident that occurs much too frequently. The CSB has investigated too many of these accidents which can be prevented by carefully monitoring for flammable vapor before and during hot work.”

This list is by no means comprehensive. In fact, after the incident in Allentown, Carl Weimer of the organization Pipeline Safety Trust was quoted in the USA Today:

Transporting natural gas by pipeline is the safest way to move that energy. Still, every nine or 10 days on average someone ends up dead or in the hospital from these pipelines. More needs to be done for safety.

And of course, pipelines are only one part of the problem.

The Marcellus Shale, the Newark Basin, and Household Income

When the US Geologic survey released their assessment of undiscovered oil and gas resources last month, it created some attention in Pennsylvania, as it raised the possibility that oil and gas companies might begin exploring areas in the southeastern portion of the state for the first time ever.  The report estimated $2.5 billion worth of gas in the southern portion of the Newark Basin at current prices.

When the legislature placed a moratorium on drilling in the formation until January 1, 2018 as impacts are studied, many observers saw this as fundamentally inconsistent with the spirit of Act 13, passed by the same legislature earlier in the year. While Act 13 established an impact fee for drilling operations and strengthened some environmental regulations, it was controversial due to all but eliminating local input on when and where gas wells and corresponding infrastructure could be built.

To many, the moratorium in southeastern Pennsylvania seems like a double standard, as many in the Commonwealth have advocated for a moratorium in the Marcellus for precisely the same reason–to assess impacts of drilling and related activity–to no avail. Why then did suburban Philadelphia get treated differently from the rest of the state?

The Marcellus Shale, the Newark Basin, and median household income by county in Pennsylvania. Please click the compass rose and double carat (^) to hide those menus. Click the blue “i” tool then any map feature for more information.

This map explores the possibility that this could be an environmental justice issue. The Newark Basin, where caution was employed, underlies the three wealthiest counties in Pennsylvania as measured by median household income according to the US Census. Obviously, correlation does not show causality, but the possibility that representatives of wealthier communities are more influential than others is an idea worth exploring.

The moratorium for the Newark Basin was inserted into the state budget at the request of Republican Senator Charles McIlhinney of Bucks County, who voted for Act 13 (known as HB 1950 until its passage). To see how other Pennsylvania senators voted on HB 1950, see the map below.

Pennsylvania senate votes on HB 1950 (subsequently known as Act 13). Click the blue “i” tool then any map feature for more information.

Word bubble using news headlines from Jackson study release

Duke Study Prompts Confusing Headlines

If you are like me and start your morning work routine by scrolling through the daily Marcellus Shale news with a good cup of coffee, then you are probably just as confused as the rest of us about the recent Duke University study results regarding shale gas drilling. Just take a look at the list below and try to interpret strictly from the news headlines what it is Nathaniel Warner, Dr. Robert Jackson, and colleagues actually found:

  • New research shows no Marcellus Shale pollution (
  • Marcellus Shale Study Shows Fluids Likely Seeping Into Pennsylvania Drinking Water (Huffington Post)
  • Rising Shale Water Complicates Fracking Debate (NPR)
  • Marcellus Brine Migration Likely Natural, Not Man-Made (Oil and Gas Online)
  • Duke study finds possible pathways from Marcellus shale to drinking water … (Akron Beacon Journal)
  • Fracking Did Not Sully Aquifers, Limited Study Finds (New York Times -blog)
  • Water contamination from shale fracking may follow natural routes (
  • Duke study: Fluids likely seeping into PA’s drinking water from Marcellus Shale (News & Observer)
  • Findings are mixed in fracking-water study (Pittsburgh Post-Gazette)
  • New study: Fluids from Marcellus Shale likely seeping into PA drinking water (
  • New research shows no Marcellus Shale pollution (The Wall Street Journal)
  • Marcellus Brine Migration Likely Natural, Not Man-Made (Duke University)
Word bubble created using Tagxedo showing news headlines from Jackson study release

No wonder this entire issue is so contentious. Not only is the science still evolving, but then you have to waft through the countless takes on what the research means. Perhaps we should take a cue from our childhood years and get the story “straight from the horse’s mouth.” E.g. try reading the official results (PDF) published in the Proceedings of the National Academy of Sciences. Even the abstract below will tell you a lot more about the implications of the results than any truncated news headline could:

The debate surrounding the safety of shale gas development in the Appalachian Basin has generated increased awareness of drinking water quality in rural communities. Concerns include the potential for migration of stray gas, metal-rich formation brines, and hydraulic fracturing and/or flowback fluids to drinking water aquifers. A critical question common to these environmental risks is the hydraulic connectivity between the shale gas formations and the overlying shallow drinking water aquifers. We present geochemical evidence from northeastern Pennsylvania showing that pathways, unrelated to recent drilling activities, exist in some locations between deep underlying formations and shallow drinking water aquifers. Integration of chemical data (Br, Cl, Na, Ba, Sr, and Li) and isotopic ratios (87Sr∕86Sr, 2H∕H, 18O∕16O, and 228Ra∕226Ra) from this and previous studies in 426 shallow groundwater samples and 83 northern Appalachian brine samples suggest that mixing relationships between shallow ground water and a deep formation brine causes groundwater salinization in some locations. The strong geochemical fingerprint in the salinized (Cl > 20 mg∕L) groundwater sampled from the Alluvium, Catskill, and Lock Haven aquifers suggests possible migration of Marcellus brine through naturally occurring pathways. The occurrences of saline water do not correlate with the location of shale-gas wells and are consistent with reported data before rapid shale-gas development in the region; however, the presence of these fluids suggests conductive pathways and specific geostructural and/or hydrodynamic regimes in northeastern Pennsylvania that are at increased risk for contamination of shallow drinking water resources, particularly by fugitive gases, because of natural hydraulic connections to deeper formations.

In all fairness, this study is very technical, so writing a catching but accurate news headline is extremely difficult. It is important to keep in mind, however, that summaries written for the lay public will often contain a piece of the translator’s perspective – like snippets of foreign code embedded in the story.

By Samantha Malone, MPH, CPH – Communications Specialist, FracTracker; DrPH Student, University of Pittsburgh Graduate School of Public Health, Environmental and Occupational Health department

Take the FracTracker Violations Quiz!

Violations issued by the Pennsylvania Department of Environmental Protection (DEP) can be found on the Compliance Report. Each violations has many columns of data, including whether it was broadly categorized as either “Administrative” or “Environmental Health and Safety” (EH&S). This is a distinction that has caused no shortage of confusion, to the point where I have argued that the distinction is actually meaningless.

But don’t take my word for it! Take the words of the DEP field agents who entered the various codes and comments. On the link below, I have made a quiz where I give you the code description or comment for ten different violations, and you use that information to decide whether you think they should be categorized as “Administrative” or “Environmental Health and Safety”.  At worst, you have a 50/50 shot at getting each question right, and you only need five out of ten to pass.

Good luck!

[Quiz expired]

Unconventional Gas Activity in Pennsylvania

Recently, the Pennsylvania Department of Environmental Protection (PADEP) Office of Oil and Gas Management changed a column on a variety of their data that they distribute. Now, instead of indicating whether or not a well is permitted or drilled into the Marcellus Shale, we are given data as to whether or not it is an unconventional well. This is a move likely designed to incorporate the Utica Shale, and perhaps other formations as well. PADEP defines unconventional wells as:

An unconventional gas well is a well that is drilled into an Unconventional formation, which is defined as a geologic shale formation below the base of the Elk Sandstone or its geologic equivalent where natural gas generally cannot be produced except by horizontal or vertical well bores stimulated by hydraulic fracturing.

Historically, of course, the lion’s share of unconventional wells in Pennsylvania have been drilled into the Marcellus Shale, although I have encountered the odd report about activity in the Utica.  Interestingly, just across the state line in Ohio, the situation is more or less reversed; evidently operators in the Buckeye State find the Utica to be more enticing than the Marcellus. In Pennsylvania, these distinctions will unfortunately be lost for us moving forward because they will be lumped together as unconventional, but really, the process is the same and the associated concerns are too. We just won’t be able to effectively compare the two black shale formations to each other in Pennsylvania.

I always feel like changes in data are a good opportunity for a retrospective. Here, perhaps for the first time ever, is a single chart with permits, violations, and drilled wells, dating back to 2005:

Obviously, June 2012 is not yet over, and the data though the 25th represents only about 83% of what we would expect for month long totals.  However, the decline in recent months is notable on all three fronts.  Let’s zoom in, so to speak, and take a look at the last 12 complete months, and add some Excel generated trend lines while we are at it:

While there is obviously significant fluctuation on a month to month basis, the negative slope of the trend lines show that these three indicators of activity for unconventional wells in Pennsylvania are all well down over a one year period.

Here is the data spatially (with violations upload pending):

Unconventional Wells and Permits in PA (large)

Notes:  At the risk of being redundant from post to post, I always like to say a few a words about how I worked with the data, just in case you want to try this at home and your graph looks a little different. The permits require a bit of preparation, because there can be multiple items listed for the same well. While that data can be valuable, it’s not really what we are looking for in this analysis. I have resolved this by using the earliest permit action for any given well API number. The drilled wells are unchanged from the original, as each well appears on the downloaded dataset exactly once. The violations are also unchanged from the original in terms of the number of records used. This results in a number of actions greater than the official DEP count of violations, which are apparently tallied by the number of violation ID’s issued. As I have mentioned elsewhere there are numerous issues with the violation dataset, and my perception is that there was a period of time in which there was a lack of uniformity in how the data were entered, which is reflected in data trail left behind. So while using all records from the data download may inflate the number of of violations, to use only the unique violation ID’s will yield a number that is too small.

Long Term Trends in PA’s Marcellus: Violations per Well

Ever since the DEP responded to FracTracker’s request for oil and gas violation data in October 2010, I have been providing periodic updates of the data in a variety of meaningful ways, such as raw violations totals and violations per amount of gas produced. But for most purposes, the best analysis has always been in terms of violations per well.

Since that time, the data have improved considerably. Not only have significant issues been addressed with both datasets, but the violations and drilled wells are now both relatively easy to access online directly from the DEP. That does not mean, however, that the available datasets are perfect or straightforward. For example, the DEP seems to count violations by the number of violation ID numbers issued, but upon closer inspection, that’s not the full story, and as a result, I prefer to use the total number of entries on the compliance report instead. The situation for permits and wells used to be almost exactly the opposite, as those reports often list multiple actions for the same well.

I have not checked the permits report lately but as I began this analysis, I was surprised to discover that the drilled wells list has been cleaned up considerably, as each unique eight digit well API number appears on the list exactly once. Now I may be the only one excited about this, but it is a notable milestone in the evolution of the data in my humble opinion,  because it removes an element of interpretation which can have a significant impact on the result of the analysis.  And as an added bonus, it makes the analysis much easier, too.

So let’s take a look at violations per well (VpW) from January 1, 2005 through June 12, 2012:

Violations per well by year in Pennsylvania’s Marcellus Shale.

Keep in mind that if we were talking about raw totals of violations and wells, it would of course be significant that we are less than half way through the year. As a ratio, however, that’s not the case, and there is no reason to expect the VpW for 2012 to change substantially up or down based on seasonality. Looking at the the trend, however, there is plenty of reason to expect the final score to be lower, as the rate of violations per well has been declining sharply in recent years. In fact, the current 2012 rate of 0.51 violations per year is less than half of the 1.14 violations per Marcellus well that we saw in 2009. Certainly, that’s an encouraging trend, if it can be attributed to changes in practices in the field, and not just changes how violations are administered, coinciding with changes in the executive branch of the state government.

One of those changes made by the new administration was an effort to route the violations process through Harrisburg. It was a move that raised considerable suspicion among some people, as it had the appearance of moving the oversight process a good bit closer to elected and appointed officials. But on the other hand, it was clear that the three DEP offices which handled oil and gas violations were not on the same page:

2010 Violations per Well by DEP issuing region.

Violations issued from the North Central Regional Office (NCRO) were roughly three times that coming from either the North West Regional Office (NWRO) or the South West Regional Office (SWRO). As the role of the regional office is supposedly diminished in determining what is and is not a violation, we will take a look at the 2012 data on a county by county basis:

Marcellus Shale violations per well (VpW) by county from 1-1-12 through 6-12-12. Counties outlined in yellow contained at least one drilled well during the period.

Compared to the map above, it seems like the strong association with VpW score and regional office affiliation is starting to break down. For those who are are curious, you can see all of the data for each county dating back to 2005 by clicking on the blue “i” tool, then any map outline.

But because the VpW scores can be so exaggerated for counties with just a handful of Marcellus Shale wells, let’s take a look at the five counties with the most wells, all of which were in the North Central Regional Office jurisdiction except for Washington, which was in the South West Regional Office:

Violations per Well for the 5 counties in PA with the most Marcellus Shale wells, as of 6-12-2012.

Last year was the first time since 2007 that Washington County did not have the lowest VpW score out of the five counties with the most wells. In this subset, trends are down across the board since 2009, and now the counties that are major players in the Marcellus are all much closer together.

When it was learned that the plan at the DEP was to have Secretary Krancer approve each Marcellus violation, the prediction that the number of violations relative to the activity of the industry would decrease was widespread. Even though the specific plan was scuttled, the expected result came to pass anyway. And yet, the validated hypothesis does not amount to proof of political meddling; the possibility that the data reflect improved practices in the field would also net the same result.

Surveying Unassessed Waters in PA

According to the Pennsylvania Fish and Boat Commission (PFBC), only 22,000 of the 86,000 miles of flowing water in PA have been sampled by biologists from their organization. As of 2011, about 12,800 miles were designated as wild trout waters. (It is hard to believe that we have so many streams to begin with!) In recent years, many groups in the Commonwealth have increased their efforts to assess these streams due to increases in potential water quality threats, such as land development and unconventional natural gas extraction. By default, unassessed streams are given the lowest classification category by the PA Department of Environmental Protection (PA DEP). It is important to prioritize streams according to their water quality, the potential for wild trout populations, and the risk posed by nearby human activities. Why trout?  Glad you asked. While there are many other ways to determine water quality, the presence of wild trout increases the streams’ water quality protection classification in PA.

A few weeks ago I spoke with an engaging gentleman from Susquehanna University, Dr. Jonathan Niles, who is working on a unique stream sampling project through the PFBC with a number of partners to do just that. Pennsylvania’s Unassessed Waters Initiative seeks to classify the 92% of streams that don’t have monitoring data about.  In 2010, PFBC partnered with two universities to survey trout populations in 30 streams each under a small grant. This work was expanded in 2011 with even more entities signing on, including Susquehanna University, and resulted in a significant increase in the number of classified streams.  The project involves entering the GPS locations of the unassessed streams and then collecting trout population data from the field. In the past two years the Unassessed Waters Initiative has surveyed 1,049 streams and documented wild trout in about 55% of those streams. Check out the progress they have made in the two maps below, the first from 2008 before sampling efforts were increased, and the second from 2012:

2008 Unassessed Waters in PA

Unassessed Waters in PA – 2008 – Unassessed streams in red, Assessed in blue

2012 Unassessed Waters

Unassessed Waters in PA – 2012 – Unassessed streams in red, Assessed in blue

In addition to the sampling protocol set forth by PFBC, Dr. Niles and his students Caleb Currens, John Panas, and Sam Silknetter collected benthic macroinvertibrate (which are PA DEP water quality indicators) and algae species data, conducted fish population estimates on every stream (not just where there was more than 5 fish of a certain species), sampled fish diets, and collected water for additional heavy metals and contaminant analysis. The preliminary fishery data from last year are currently being reviewed by the PADEP.

Some of the Initative’s efforts have focused on the quality of streams near shale gas drilling operations, especially due to the risk that erosion and sedimentation poses to trout’s habitat. Dr. Niles feels that the data collected from initiatives like this one provide valuable operating insight for development and natural gas companies, as sensitive areas can be avoided by companies – saving them time and money.

With funding from the National Fish and Wildlife Foundation, the Iinitiative has been funded again this year. Dr. Niles’ team is contracted to assess 20 streams in Loyalstock that were previously unassessed. An additional 40 streams will be assessed by Dr. Niles’ team elsewhere in PA. While they have made extraordinary progress, there is still much work to be done. What does a project with such a broad geographic scope like this one cost? In addition to travel and salary costs, each benthic macroinvertebrate sample runs about $200-250 to analyze in a lab. It is likely that this year alone there will be at least 60 samples collected by Dr. Niles’ team, if not more. The financial cost of conducting this kind of research may seem high, but the failure to do so could cost Pennsylvanians much more. It is our hope, here at FracTracker, to keep up-to-date with the Unassessed Waters Initiative as the teams go out this year. Check back soon for more information, or contact us if you would like to get involved with either the sampling or funding of this initiative:

Below are photos of Dr. Niles’ assessment team taken during their field sampling trips.

In addition to the Fish and Boat’s own crews, the following 15 groups are partners for this year’s Unassessed Waters Initiative:

  • Penn State University
  • California University of Pennsylvania
  • Susquehanna University
  • Clarion University
  • Lycoming College
  • Kings College
  • Keystone College
  • Juniata College
  • Allegheny College
  • Mansfield University
  • Lock Haven University
  • Duquesne University
  • Loyalsock Creek Watershed Association
  • Western Pennsylvania Conservancy
  • Trout Unlimited – Eastern Abandoned Mines program

By Samantha Malone, MPH, CPH – Communications Specialist, FracTracker; and DrPH Student, University of Pittsburgh, Graduate School of Public Health, Environmental and Occupational Health department. (email)

Special thanks to Jon Niles (Susquehanna University) and Bob Weber (PA Fish and Boat Commission) for their contributions to this article and efforts in the field!

US Natural Gas Consumption Patterns

Ever wonder what we do with all the gas we’ve been producing?

In 2011, the United States consumed about 24.4 trillion cubic feet (Tcf) of natural gas, a 2.4 percent increase from 2010. From that total, just over 2 Tcf was used to fuel operations necessary to get the product to market, including production and recovery efforts as well as transportation of the product through pipelines. The remainder of the natural gas, about 22.3 Tcf, made its way to consumers.

Residential, commercial, and vehicle fuel usage have all been relatively flat over the past dozen years, save for a slight dip between 2005 and 2007, when the wellhead price of gas was around $6 to $8 per thousand cubic feet (Mcf).  On the other hand, the amount used for electric power has increased 46 percent, while industrial usage has fallen 17 percent over the same period of time.

Much of this variation may be due to external factors, such as changes in industrial production or the fact that price of coal roughly doubled over the same time frame.  And as a whole, the United States must have made significant progress in increased residential efficiency, because although there are 32 million more people now than in 2000, the residential sector actually saw a slight decrease in overall natural gas consumption.

But has the recent surge in domestic drilling activity reduced our dependence of foreign oil? Although oil  imports were about one third less in 2011 than the 2005 peak, many experts think that is due to a combination of high prices (due mainly to Asian demand) and the less-than-robust US economy making fill-ups at the pump more painful, and therefore, less frequent.

Consider that in the United States, the way we use natural gas and petroleum are really quite different. While many developing nations have millions of natural gas vehicles, the United States had fewer than 120,000 compressed natural gas (CNG) and liquefied natural gas (LNG) powered vehicles in 2010, explaining why vehicle fuel is barely a factor in the chart above.  Compare that to petroleum, below, where gasoline, diesel, and jet fuel account for nearly three quarters of all petroleum usage:

Consumers can hardly be faulted for not buying natural gas powered vehicles: according to an interactive map on the US Department of Energy site, there are exactly seven CNG filling stations within a 100 mile radius of downtown Pittsburgh, and no LNG stations within that area.

So natural gas is not yet a major factor in replacing our primary use of petroleum, which is really to move people and things around. That could change though, if production remains high and prices remain low. That last part might be what has prevented a movement toward natural gas vehicles: currently, the price of CNG is about $2.13 per gallon equivalent, but if gas prices go back up to around $8 per Mcf (from the current $2.78) then gas may not seem like such a good deal again. One can only presume that fluctuations such as those factor into why the investments for CNG and LNG infrastructure is so limited.

A Look at the Haynesville Shale

The Haynesville Shale is a gas rich formation that underlies parts of Louisiana, Texas, and Arkansas. Technically a mudstone, it was deposited in the Upper Jurassic epoch, about 150 million years ago. As the chart below demonstrates, the Haynesville Shale is a major player in terms of shale gas extraction in the United States:

US shale gas production by formation. Source: US Energy Information Administration

Unfortunately, the exact production values were not released along with this chart, but the Haynesville is clearly competitive with both the Barnett Shale and the Marcellus Shale, and may in fact outproduce both of them. So let’s take a look at the Haynesville, and with it, a look at the data distribution of the three states where it is located.

Haynesville Shale activity. Please click the grey compass rose and double carat (^) to hide those menus. For more information, click the blue “i” tool, then any map feature.


Arkansas has 64 wells drilled into the Haynesville formation, only a handful of which were drilled in the 21st Century. In fact, many of the wells listed are quite old, dating to the 1940’s. Taken together, the small number of wells and the age of those wells makes it clear that Arkansas is not much of a factor in the total output of the Haynesville Shale. In fact, the Haynesville wells in Louisiana stop well south of the border with the Natural State.

In terms of their data, the Arkansas Oil and Gas Commission provides access to data that many other states do not, including well stimulation data. Unfortunately, the data export feature seems to be a bit buggy, but luckily, the Haynesville data is practically unchanged from when I downloaded the data in 2010.  Much of the dataset has location information in decimal degrees–the standard latitude and longitude format of the digital age.  Many of the wells are only available in the archaic and arcane Public Land Survey System, which is still the legal way to describe locations in many western states.

In all likelihood, the Pelican State contains the largest share of recoverable gas from the Haynesville formation.  This assertion is based on activity rather than production:  2,400 Haynesville wells have been permitted since 2006, of which 2,194 have been spudded, and 1,985 have been completed.

The data released by Louisiana’s Department of Natural Resources has a certain elegance to it, in that the dates for permits, spuds, and original well completion can all be found on the same spreadsheet. What’s more, the same dataset contains location information in the preferred decimal degree format for the top and bottom of the wells, meaning that it is possible to map the lateral portion.

With almost 100,000 producing gas wells in 2010, Texas accounts for one fifth of all such wells in the United States. The heart of the Haynesville Shale continues into the Lone Star State from the western border of Louisiana.

And yet, there are no wells in Texas on the map above. Why? I can tell you that as of the writing of this, there are 816 Haynesville wells in Texas, according to the well search tool of the agency which oversees the oil and gas industry, the Railroad Commission of Texas (RRC). I can quickly find production values, and even the lease number, and in other queries, there are even scans of original paperwork available online, digitally, and free of charge. What they don’t provide is location data. This omission seems to be consistent with the following statement on the RRC site:

The on-line research queries are intended for use by individuals needing specific information from the Railroad Commission. The use of automated tools to retrieve volumes of data can cause severe degradation of the Railroad Commission’s systems. If the query system detects that data is being retrieved using an automated tool, the Commission will end the session for that user.

The first sentence contains the philosophy, while the second statement frankly comes across as a lame excuse in 2012, especially for a state with such substantial activity and a healthy severance tax. It seems inconsistent to dedicate so much server space to scans of images without informing you where the location is–it’s a convoluted approach to data transparency, if ever there were one.

Perhaps my interest in mapping is a source of bias, but to me, the location of the well is the single most important data point. How can you determine what the potential impact of a well would be if you don’t even know where it is?

Pennsylvania Marcellus Fines Data

Usually, I try to let readers know the state of affairs related to oil and gas extraction by taking a hard look at publicly available data.  Sometimes, however, it seems like the simplest questions have an answer that starts off with, “Well, it’s complicated…”  Such is the case when it comes to fines issued by the the Pennsylvania Department of Environmental Protection’s (PADEP) Office of Oil and Gas Management.  Luckily  PADEP releases data about fines issued to operators in its compliance report, but unfortunately, it can be confusing to interpret.  Let’s take a closer look:

The first point of confusion is the compliance report itself.  Specifically, there are more rows of data than there are violations, as counted by PADEP.  The answer seems to be that PADEP counts the number of unique violation ID numbers, however sometimes (but not always), the same Violation ID will be used for multiple rows of data on the violation report.  When I downloaded Marcellus violation data from January 1, 2005 through May 2, 2012, there were 4,293 rows of violation data, but only 3,689 unique violation ID’s.

Distribution of Violation ID frequency on the PADEP Marcellus Shale compliance report, January 1, 2005 through May 2, 2012.

If a violations are counted by unique Violation ID numbers, what then are we to make of the other 604 items on the list? They all have violation numbers, but share them with between one and three other incidents. My perception is that this is one of those decisions made in the field that has unanticipated consequences with respect to database maintenance. That is just a guess though–I’ve contacted PADEP for clarification on this point, and will be sure to relay that information when I receive it.

The second confusing aspect of the fines data is fairly similar in nature, in that identical fine amounts will often appear for multiple violations (or rows of violation data, as the case may be).  This is an ostensibly reasonable thing to do; if PADEP can deal with a suite of related violations all at once, why not do so?  But it does beg the question of whether the full fine is posted for each item on the compliance report, or whether it has been prorated between them.

I believe the former case to be correct.  Take for example, the recent announcement of a fine issued by PADEP to Ultra Resources for improper storage of flowback water at a Potter County site.  The announcement mentions a $40,000 fine, but the data reflects three fines assessed to Ultra for that amount on March 23, 2012 for three incidents with unique Violation ID numbers.

The implication is that if you go through the dataset and add up the value of all the fines, the result will almost certainly be wildly inflated.  The same is true for the number of fines that have been assessed.   However, there is something linking the three records together in the Ultra example:  they all share the same Enforcement ID number.  So perhaps that is the key?  Let’s take a look at the total number and value of Marcellus fines assessed, with the data organized in three different ways:

Aggregated number and value of Marcellus related fines in PA from 1-1-2005 to 5-2-2012, by method.

Given that the three Ultra violations in the example above all had unique Violation ID numbers but shared the same Enforcement ID, my expectation was that aggregating the data by unique Enforcement IDs would yield the smallest (and most accurate) statewide totals.  Clearly, that hypothesis needs to be relegated to the scrap heap, based on the table above.

And to be honest, I don’t have a better hypothesis on deck.  I have also asked PADEP for clarity on this point, and will be happy to share that information when I receive a reply.  But for now, I’m not even sure if it is possible to tease the correct answers from the data that have been provided.  Which is a shame, because if we knew a reliable methodology for doing so, it would be possible to explore the topic in much more interesting detail, finding answers for questions like:  Which company gets fined the most?  What’s the ratio of violations to fines assessed?  How many days pass between a violation being issued and a fine?  (For this last one, I can tell you that the maximum amount of time so far is 755 days–I just can’t provide a reliable distribution of the results).

I like to give the DEP credit where it is due:  they are making tremendous progress in their dissemination of oil and gas data.  Two years ago, there were no compliance, production, or waste reports.  Drilled well data was available, but much of it didn’t have location data, and you had to copy and paste from web tables to a spreadsheet, which didn’t always work very well.  And some of the location data for permits were miles away from the actual well site with the corresponding API number.  PADEP has come a very long way in the reliability and accessibility of their oil and gas data.  Here’s hoping that trend continues.