Amid Oil and Gas Boom, Colorado Continues Role as Earthquake Lab

- by Kevin Simp­son, August  31, 2014,  The Den­ver Post

From the liv­ing room chair where he sat read­ing around half past 9 on a May evening, Ron Bak­er heard the boom and felt his cen­tu­ry-old Gree­ley farm­house shud­der, send­ing a menagerie of plas­tic hors­es top­pling from a bed­room shelf.

He stepped out the back door and aimed a flash­light at the thick, ancient cot­ton­wood that leans over the roof, expect­ing to reveal a snapped limb as the cul­prit. But he cir­cled the house and found noth­ing amiss.

About a half-mile down the coun­ty road, Judy Dunn had been sit­ting in bed watch­ing TV when she felt her brick ranch house shake and heard the win­dows rat­tle, mak­ing her won­der if an oil or gas well had blown.

A few miles away in the city, Gail Jack­son joined neigh­bors spilling out into the street, won­der­ing if a plane crash had trig­gered the big bang and sud­den vibra­tion that dis­si­pat­ed as quick­ly as it arrived.

All over, phones rang and neigh­bors com­pared notes as the mys­tery unrav­eled: Weld Coun­ty had felt the tremors of a magnitude‑3.2 earth­quake — jar­ring but accom­pa­nied by lit­tle, if any, damage.

In an area pep­pered with wells pulling ener­gy resources from below ground — and many pump­ing waste­water from the process back into it through injec­tion wells — an old ques­tion resur­faced: Could the same geo­log­i­cal tin­ker­ing that has revved a for­mi­da­ble eco­nom­ic engine also trig­ger poten­tial­ly dam­ag­ing earthquakes?

“I knew there had been spec­u­la­tion around injec­tion wells caus­ing seis­mic activ­i­ty,” Bak­er says. “This sort of con­firmed what I’d been reading.”

Spec­u­la­tion has turned to full-on inves­ti­ga­tion as researchers from the Uni­ver­si­ty of Col­orado jumped at the chance to gath­er data — and with­in days had set up a net­work of seis­mome­ters sur­round­ing the esti­mat­ed epicenter.

State reg­u­la­tors even­tu­al­ly zeroed in on one high-vol­ume injec­tion well and had its oper­a­tor shut off the flow for 3½ weeks before resum­ing activ­i­ty on a grad­ual basis, while the CU sci­en­tists track seis­mic activ­i­ty nearby.

The unex­pect­ed oppor­tu­ni­ty revives the con­cept of “induced seis­mic­i­ty” explored in Den­ver more than half a cen­tu­ry ago at the Rocky Moun­tain Arse­nal chem­i­cal weapons plant and then in oil and gas fields of west­ern Col­orado into the 1970s.

The Bureau of Recla­ma­tion has tracked induced seis­mic­i­ty since the 1990s in a riv­er desali­na­tion project in the Para­dox Val­ley, but the issue large­ly slipped under the radar until the indus­try boom of the past sev­er­al years.

The Gree­ley quake, in a region not known for exten­sive seis­mic activ­i­ty, came on the heels of research out of Okla­homa, a state also under­go­ing inten­sive oil and gas extrac­tion and waste­water injec­tion. That study linked a stun­ning spike in earth­quake activ­i­ty to the pres­sur­ized flu­ids pumped far under­ground — where, sci­en­tists say, they migrat­ed to and essen­tial­ly lubri­cat­ed exist­ing faults.

But while inter­est has ramped up in Col­orado and else­where, the issue remains far from set­tled. The wells, long regard­ed as an envi­ron­men­tal­ly respon­si­ble way to dis­pose of waste­water, have pierced the land­scape for years, most­ly with­out seis­mic drama.

And so indus­try offi­cials and oth­ers express skep­ti­cism about any link, even as they coop­er­ate, to vary­ing degrees, with researchers. Mean­while, 334 dis­pos­al injec­tion wells in Col­orado pump waste­water deep under­ground, with com­pa­nies cur­rent­ly seek­ing approval for 35 more — like one near Roy Wardel­l’s ranch in Platteville.

Wardell, 73, did­n’t feel the earth shake on his 3,500-acre Weld Coun­ty spread. But he’d been say­ing for a while that too many injec­tion wells are con­cen­trat­ed near his prop­er­ty — five of them with­in a 2‑mile radius.

He opposed a per­mit for the newest pro­posed well after read­ing about seis­mic­i­ty con­cerns in Okla­homa, Ohio and Texas. And then Gree­ley was added to the list.

“I felt a lit­tle bit like a prophet, already voic­ing that con­cern before it hap­pened,” says Wardell, who in addi­tion to rais­ing pure­bred Angus, draws oil and gas rev­enue from wells on his prop­er­ty. “You have to bal­ance. Just because you might have some good income, you don’t want to throw away your health and safe­ty and land. I felt the earth­quake thing is much big­ger than my place.”

He thought his con­cerns were heard by the Col­orado Oil and Gas Con­ser­va­tion Com­mis­sion, the pan­el charged with both pro­mot­ing and reg­u­lat­ing the indus­try. But he was not sur­prised that the per­mit was granted.

“Col­orado is ground zero”

Human-caused events — from mine explo­sions to nuclear tests to flu­id injec­tion — have rat­tled the state over decades, but it’s the last cat­e­go­ry that has par­tic­u­lar res­o­nance in these oil and gas boom times.

“Col­orado is ground zero for research on induced earth­quakes caused by injec­tion,” says Bob Kirkham, a con­sult­ing geol­o­gist who for­mer­ly worked for the Col­orado Geo­log­i­cal Sur­vey. “Yet still there are hun­dreds of injec­tion wells that aren’t caus­ing quakes. There must be some­thing geo­log­i­cal­ly that’s going on respon­si­ble for some caus­ing earth­quakes and some not, but we don’t know the answer at this time.”

When the Army began pump­ing liq­uid waste from its Rocky Moun­tain Arse­nal chem­i­cal weapons plant in 1962, thou­sands of small earth­quakes ensued, and a cou­ple exceed­ed mag­ni­tude 5.0 — with one caus­ing $1 mil­lion dam­age to near­by Com­merce City.

When geol­o­gist David Evans took note of the seis­mic activ­i­ty and even­tu­al­ly found a cor­re­la­tion between the num­ber of quakes and the vol­ume of inject­ed liq­uid at a giv­en time, the Army denied any con­nec­tion. Oth­er geol­o­gists also were skep­ti­cal about his find­ings until the U.S. Geo­log­i­cal Sur­vey launched its own study, which ulti­mate­ly vin­di­cat­ed Evans.

That dis­cov­ery led to the USGS team­ing with the oil indus­try in the Range­ly oil field in north­west­ern Col­orado to deter­mine if water injec­tion, which seemed to be trig­ger­ing fre­quent minor quakes, could effec­tive­ly func­tion as a seis­mic switch.

When injec­tion stopped, dozens of earth­quakes per day fell to few­er than 10. When it resumed, so did the fre­quen­cy of seis­mic activ­i­ty. The exper­i­ment was repeat­ed sev­er­al times into the ear­ly 1970s.

In the Para­dox Val­ley, efforts to reduce salt con­tent in the Dolores and Col­orado rivers led to anoth­er close­ly watched exper­i­ment: The Bureau of Recla­ma­tion inter­cept­ed salty ground­wa­ter before it hit the Dolores and put it in evap­o­ra­tive ponds and, start­ing in 1996, inject­ed it underground.

The bureau had copi­ous pre-injec­tion seis­mic data for com­par­i­son — there was­n’t much going on — and set up micro­seis­mome­ters to mon­i­tor data for years. Thou­sands of earth­quakes have been induced, most so small they could not be felt.

“When they start­ed to inject, they start­ed hav­ing quakes right at the bot­tom of the well,” Kirkham says. “You could see where they start­ed, how they migrat­ed away from the well, found a fault zone and start­ed using it as a place for epicenters.”

After a magnitude‑4.3 quake in 2000, they dialed back the injec­tion rate with the desired result — no more activ­i­ty over 4.0.

“If we know where that thresh­old is, you can con­tin­ue to inject and not cause prob­lems,” Kirkham notes. “It’s impor­tant to iden­ti­fy why cer­tain areas are prone to injec­tion-induced quakes and some aren’t.”

Debate over induced seis­mic­i­ty recent­ly has played out in south­ern Col­orado, where earth­quake swarms near Trinidad — includ­ing a magnitude‑5.3 quake in 2011 — have pro­duced dif­fer­ences of opin­ion over pos­si­ble man-made or nat­ur­al caus­es. The USGS con­tends that at least a major­i­ty of the quakes are relat­ed to waste­water injec­tion, while indus­try peo­ple and the Col­orado Geo­log­i­cal Sur­vey counter that such an assess­ment is premature.

USGS geo­physi­cist Justin Rubin­stein co-authored the study link­ing seis­mic activ­i­ty near Trinidad and in Okla­homa to injec­tion wells and now works in the agen­cy’s Induced Seis­mic­i­ty Project.

He would love to see anoth­er study like the Range­ly exper­i­ment in the ’60s, but notes that it would require a sel­dom-seen part­ner­ship between inde­pen­dent researchers and industry.

In the mean­time, one of the main goals of the USGS project is to more ful­ly grasp the haz­ards asso­ci­at­ed with injection.

“We can’t say there’s no risk of there being sig­nif­i­cant dam­age and loss,” Rubin­stein says. “Try­ing to quan­ti­fy that is an area of active research — one of the most impor­tant things we’re doing now.”

At mag­ni­tude 6.6, Col­orado’s largest record­ed earth­quake shook a much less pop­u­lous state in 1882. It took more than 100 years for sci­en­tists to dial in its nat­ur­al ori­gin. Kirkham and a col­league placed it about 10 miles north of Estes Park in a 1986 study that was lat­er con­firmed by USGS research.

Even today, experts note that there’s a good deal we don’t know about what lies beneath the state’s diverse landscape.

“We’re at the begin­ning stages of try­ing to under­stand the seis­mic activ­i­ty in the state, even the faults we have,” says Matt Mor­gan, senior geol­o­gist at the Col­orado Geo­log­i­cal Sur­vey. “There are about 90 faults that we know of with poten­tial of slip­ping. But most of the events may occur on fea­tures that are unmapped.”

One pro­pos­al in search of fund­ing would put dozens more per­ma­nent seis­mome­ters around the state, with data fed to the Nation­al Earth­quake Infor­ma­tion Cen­ter at the USGS. With­in a cou­ple of years, Mor­gan esti­mates, geol­o­gists would gain a bet­ter under­stand­ing of the forces at work across the state.

Among oth­er things, it could assist in the cre­ation of the seis­mic haz­ard maps pro­duced by the USGS every six years. The most recent edi­tion, released this sum­mer, did­n’t account for quakes that may have been induced, prompt­ing some experts to ques­tion whether Col­orado’s risk has been underestimated.

Mark Petersen, the nation­al region­al coor­di­na­tor for the USGS Earth­quake Haz­ards Pro­gram, says that while the maps side­stepped the issue of poten­tial­ly induced earth­quakes by ignor­ing them in this edi­tion, plans are under­way to fac­tor them in.

Matt Lep­ore, direc­tor of the COGCC, notes that under­ground injec­tion is the best dis­pos­al option avail­able, dis­tin­guish­ing the process from recy­cling, which also has been on the rise. As the increased oil and gas activ­i­ty has pro­duced greater demand for dis­pos­al wells, it also car­ries the need to eval­u­ate the capac­i­ty of the system.

“We need to get every­thing lined up in that respect,” Lep­ore says, “and make sure that there’s enough capac­i­ty to oper­ate wells the way they’re per­mit­ted and designed to oper­ate, which in our view keeps those risks at a rea­son­able level.”

Strong opin­ions

When William Yeck asked a landown­er east of the Gree­ley air­port for per­mis­sion to put a small seis­mome­ter sta­tion on his prop­er­ty, he was told in no uncer­tain terms: Frack­ing does­n’t cause earthquakes.

Opin­ions tend to be firm­ly held — on both sides of the frack­ing issue — in an area where oil and gas pro­duc­tion has become a major eco­nom­ic force as well as an envi­ron­men­tal concern.

And while frack­ing and waste­water injec­tion wells go hand in hand, experts dis­tin­guish between the process of hydraulic frac­tur­ing to extract oil and gas and the dis­pos­al of the pro­duced water pumped deep under­ground — ide­al­ly to dis­perse into porous rock.

It’s this lat­ter process that has become the object of scruti­ny with regard to seis­mic activity.

Yeck, a doc­tor­al can­di­date at CU, explained the ratio­nale behind the study of the recent Gree­ley earth­quake: It occurred in an unusu­al spot, and one hypoth­e­sis holds that the seis­mic activ­i­ty may have been induced. By sta­tion­ing equip­ment in the area to mon­i­tor activ­i­ty, researchers could test the hypothesis.

The landown­er con­sent­ed, and Yeck buried a seis­mome­ter about a foot deep in farm­land at the edge of a corn­field that looks out onto a flat expanse where work­ers attend to two drilling rigs.

A short dis­tance away, trucks rum­ble into a waste­water facil­i­ty that includes a well des­ig­nat­ed C4A — the injec­tion well whose activ­i­ty and prox­im­i­ty to the earth­quake’s epi­cen­ter have made it a focal point of observation.

One August morn­ing, Yeck parks near the portable sta­tion, clears some tall weeds that have sprout­ed around it and opens the green plas­tic, oblong box out­fit­ted with a solar pan­el for pow­er. He checks the data log­ger that acquires infor­ma­tion from the seis­mome­ter, con­nects an iPod Touch to make sure the unit’s GPS set­tings and data con­fig­u­ra­tions are cor­rect, replaces the cur­rent data disk with a new one and does a “stomp test” — pound­ing his foot on the ground to make sure the instru­ment is read­ing ground vibrations.

On this unit, he also checks the cell­phone modem that trans­mits real-time data that can be mon­i­tored by the oil and gas com­mis­sion, the well oper­a­tor and con­sult­ing com­pa­nies, as well as the pub­lic. Only two sta­tions among the six scat­tered around the quake area have that capa­bil­i­ty. Oth­er data are retrieved at rough­ly two-week intervals.

The project, head­ed by CU geo­physics pro­fes­sor Anne Shee­han, con­tin­ues a revival of inter­est in the field of induced seis­mic­i­ty that has accom­pa­nied the surge in oil and gas production.

Before the Gree­ley quake, the most stun­ning research this sum­mer had come from Okla­homa, where researchers — includ­ing two of Shee­han’s col­leagues at CU — attrib­uted a mas­sive increase in seis­mic activ­i­ty to injec­tion wells, claim­ing that the pres­sur­ized waste­water seeped into faults and caused a spike in activity.

Okla­homa had aver­aged about two quakes of mag­ni­tude 3.0 or high­er before 2008. In the first four months of this year, the U.S. Geo­log­i­cal Sur­vey record­ed 145.

After the Gree­ley quake, Shee­han saw an oppor­tu­ni­ty to gath­er data relat­ed to seis­mic activ­i­ty in the vicin­i­ty of the C4A well run by NGL Water Solutions.

“In oth­er places like Okla­homa, seis­mic­i­ty is near­ly out of con­trol,” Shee­han says. “If that’s poten­tial­ly what’s hap­pen­ing in Col­orado, we want to nip it in the bud. If it’s relat­ed to a well, let’s fig­ure out the para­me­ters they can oper­ate at safely.”

When the COGCC asked NGL to stop injec­tion after a magnitude‑2.5 quake June 23, researchers had the oppor­tu­ni­ty to look at seis­mic activ­i­ty while the well was dor­mant and lat­er as reg­u­la­tors autho­rized the com­pa­ny to resume injec­tion at low­er but grad­u­al­ly increas­ing rates.

The com­mis­sion gave the OK to resum­ing injec­tion of waste­water at C4A on July 18. But instead of 17,000 bar­rels per day, NGL could inject only 5,000 bar­rels per day for 20 days. On Aug. 7, that vol­ume increased to 7,500 bar­rels per day — all at the same max­i­mum injec­tion pressure.

If all con­tin­ues to go well, the com­mis­sion antic­i­pates autho­riz­ing NGL to push the vol­ume to 10,000 bar­rels per day, with room for lat­er nego­ti­a­tion if the rock proves porous enough to accept waste­water easily.

The CU study will con­tin­ue to mea­sure seis­mic activ­i­ty in the area until win­ter, but so far it has detect­ed hun­dreds of small earth­quakes of a mag­ni­tude gen­er­al­ly below the “felt range” that cor­re­late to the loca­tion of the well.

Only four have been record­ed over mag­ni­tude 2.0, most recent­ly a magnitude‑2.1 on Aug. 13. The COGCC has con­tin­ued to mon­i­tor activ­i­ty and could halt injec­tion again if quakes of 2.5 or above occur with­in a 2.5‑mile radius of the well.

But while sig­nif­i­cant activ­i­ty has been qui­et recent­ly, Yeck notes that induced seis­mic­i­ty in part depends on the core pres­sure in the rock where injec­tion has been tak­ing place. That pres­sure would dis­si­pate when injec­tion is stopped, but once injec­tion is resumed, it could take a while to repres­sur­ize the system.

Since the shut­down and restart, seis­mic­i­ty rates have not increased. One fac­tor could be action tak­en by NGL to seal part of the well with a cement cas­ing to alter the path of inject­ed water away from so-called “base­ment” rock.

“At the moment, it looks like it may have worked, because the earth­quakes are very small,” Shee­han says. But it could also be that in two months’ time, there’s going to be enough flu­id that earth­quakes will start up again.”

In the mean­time, the COGC­C’s Lep­ore says the com­mis­sion’s stance on any con­nec­tion between the Gree­ley quake and the injec­tion well remains in the “not defin­i­tive­ly caused by” category.

“I think we have respond­ed in a way that we feel is appro­pri­ate, and we’re hap­py with the results so far,” Lep­ore said. “We’ll stay on the path we’re on unless new data emerges.”

NGL, which owns 10 wells at sev­en facil­i­ties in east­ern Col­orado’s Den­ver-Jules­burg Basin with plans to add sev­en more wells at five new Weld Coun­ty facil­i­ties in the next 18 months, has been a will­ing part­ner to the study. And while NGL senior vice pres­i­dent Doug White says he’ll leave the sci­ence to the experts, he stands by the prac­tices that have worked for his company.

“We have wells that have been inject­ing water for 20 years with no issues,” White says. “Until we see defin­i­tive sci­en­tif­ic data that con­nects the two, we’re busi­ness as usu­al based on how we’re reg­u­lat­ed by the oil and gas commission.”

Anadarko Petro­le­um Corp., which uses both its own wells and oth­er com­mer­cial wells for dis­pos­al, point­ed out that injec­tion wells are reg­u­lat­ed, test­ed fre­quent­ly and have lim­its on their vol­ume and pressure.

“Though some stud­ies have shown injec­tion can cause induced seis­mic­i­ty, the mag­ni­tude is sim­i­lar to a train pass­ing by well below the sur­face of the earth,” Kor­by Brack­en, direc­tor of health, safe­ty and envi­ron­ment for Anadarko Rock­ies divi­sion, wrote in an e‑mail. “…With ade­quate reg­u­la­tion already in place, water dis­pos­al wells have been uti­lized for many years, by many indus­tries, safe­ly and effectively.”

The CU researchers remain cau­tious in draw­ing any con­clu­sions until the study is com­plet­ed, peer reviewed and pub­lished. Even­tu­al­ly, the CU hydrol­o­gists who aid­ed the Okla­homa study will use the data, as well as oth­er infor­ma­tion about the per­me­abil­i­ty of the geo­log­ic for­ma­tions in the area, to run com­put­er mod­els show­ing where the inject­ed flu­ids could migrate over time.

Mean­while, Col­oradans wait for defin­i­tive word on under­ground rum­blings that, depend­ing on point of view, mer­it a shrug of the shoul­ders or instill a sense of foreboding.

“Of course it’s a con­cern, but by and large we believe it’s one that can be man­aged,” says COGC­C’s Lep­ore. “You know when you go out there, there are risks. But you try to be pre­pared and under­stand those risks and antic­i­pate and avoid them. How we han­dle the C4A well reflects how we han­dle what we believe is a man­age­able situation. ”

Kevin Simp­son: 303–954-1739, ksimpson@denverpost.com or twitter.com/ksimpsondp