研究顯示注入到地下的流體可以在相當遠的地方誘發地震

原文網址：https://now.tufts.edu/news-releases/study-suggests-earthquakes-are-triggered-well-beyond-fluid-injection-zones

研究顯示注入到地下的流體可以在相當遠的地方誘發地震

科學家利用電腦模擬和現地實驗得到的數據，得出注入到地下的流體和群震間的新關聯

Mike Silver

探勘油氣的過程中，會把流體注入地下來壓裂地層或者處置廢水。美國塔夫茨大學的研究人員利用現地實驗和斷層模擬的數據，發現注入地下的流體可以迅速在流體擴散的範圍之外，產生顯著的地震活動。之前的研究已經知道注入深層流體 (超過一公里深)和地震活動增加有關，但一般認為影響範圍只侷限在液體擴散到的區域。然而今日發表在期刊《科學》(Science)的研究中，試驗結果強烈支持了以下理論：注入流體可以讓早已存在的斷層破裂網路緩緩滑移，就像推骨牌一般，到了很遠的地方還是有機會發生地震並造成災害。

奧克拉荷馬州人為引發的地震比加州的天然地震還要活躍。藍色框框標記了有許多人為引發地震的區域。來源：美國地質調查局

這項結果可以解釋美國某些地區人為產生地震的頻率為何超過了天然地震的熱點。

研究也證明了透過實際在現地進行的實驗，來發展並測試更加精確的斷層行為模型是可行的做法。目前對於斷層物理機制的理解，大部分是來自於在實驗室進行的實驗，用的樣品只有一公尺甚至更小。然而，地震時斷層的實際破裂規模相當巨大。目前只能從遠處觀察斷層在如此大尺度之下的破裂，但以此得到的斷層行為物理參數，在模擬人類對斷層產生的效應時，不過是非常粗糙的估計值。最近地震科學界開始投入資源進行現地規模的注水實驗，目標是統合兩種尺度下的觀察結果，並瞭解自然條件下的斷層性質。

之前法國艾克斯―馬賽大學和尼斯大學組成的研究團隊在法國進行了現地注水實驗，他們測量了斷層的增壓情形、位移量、滑動量.......等物理參數，而本研究再把這些參數匯入斷層滑移模型當中。塔夫特大學研究人員的分析結果是目前最為有力的推論，指出因流體而活化的斷層滑移傳遞速度，可以迅速超過流體在地下的擴散速度。

「在建立精確的地震災害數值模型時，一個相當重大的阻礙是我們缺少斷層在自然條件下的性質。」研究主要作者Pathikrit Bhattacharya表示。他之前是塔夫特大學工學院土木與環境工程學系的博士後研究員，現為印度理工學院(位在布巴尼斯瓦爾)地球海洋與氣候科學院的助理教授。「這些結果證明了如果能取得這類現地觀測數據，我們可以更加深入地瞭解斷層的力學性質，並迫使我們重新思考斷層可能造成多大的災害。」

流體引發的地震災害在美國越來越受公眾關注。奧克拉荷馬州的油氣探勘活動相當頻繁，一般認為人為製造的地震導致奧克拉荷馬州變成美國地震最活躍的地區，甚至超越了加州。「驚人的是，現今某些地區的人為地震頻率竟然超出了天然地震活動的熱點，像是加州南部。」研究共同作者Robert C. Viesca表示。他是塔夫特大學工學院土木與環境工程學系的副教授，也是Bhattacharya的博士後指導教授。「我們的結果證明了把流體注入到地下深處可能引發的後果，也提供了有力的工具，讓未來的油氣探勘可以評估誘發地震的傳遞及風險。」

大多數水力破裂法引發的地震規模都小於芮氏3.0，不需要擔心會造成災害或危及生命。然而，把探勘產生的廢棄物注入地下深處可能會影響地下深部的大型斷層，由於它們處於壓力之下，容易受液體的影響而滑動。因此，把廢水注入比一公里還深的鑽井，可能會產生較大型的有感地震並造成危害。

根據美國地質調查局表示，科學文獻中由流體注入引發的地震，規模最大的是2016年9月在奧克拉荷馬州中部的規模5.8地震。其他四次由流體注入引發且規模大於5.0的地震同樣發生在奧克拉荷馬州；阿肯色州、科羅拉多州、堪薩斯州、德州也曾經因為流體注入而發生規模4.5到5.0的地震。

Study suggests earthquakes are triggered well beyond fluid injection zones

Computer model and field experiment data suggest a new link between subsurface injections and earthquake swarms

Using data from field experiments and modeling of ground faults, researchers at Tufts University have discovered that the practice of subsurface fluid injection used in ‘fracking’ and wastewater disposal for oil and gas exploration could cause significant, rapidly spreading earthquake activity beyond the fluid diffusion zone. Deep fluid injections -- greater than one kilometer deep -- are known to be associated with enhanced seismic activity—often thought to be limited to the areas of fluid diffusion. Yet the study, published today in the journal Science, tests and strongly supports the hypothesis that fluid injections are causing potentially damaging earthquakes further afield by the slow slip of pre-existing fault fracture networks, in domino-like fashion.

The results account for the observation that the frequency of man-made earthquakes in some regions of the country surpass natural earthquake hotspots.

The study also represents a proof of concept in developing and testing more accurate models of fault behavior using actual experiments in the field. Much of our current understanding about the physics of geological faults is derived from laboratory experiments conducted at sample length scales of a meter or less. However, earthquakes and fault rupture occur over vastly larger scales. Observations of fault rupture at these larger scales are currently made remotely and provide poor estimates of the physical parameters of fault behavior that would be used to develop a model of man-made effects. More recently, the earthquake science community has put resources behind field-scale injection experiments to bridge the scale gap and understand fault behavior in its natural habitat.

The researchers used data from these experimental field injections, previously conducted in France and led by a team of researchers based at the University of Aix-Marseille and the University of Nice Sophia-Antipolis. The experiments measured fault pressurization and displacement, slippage and other parameters that are fed into the fault-slip model used in the current study. The Tufts researchers’ analysis provides the most robust inference to date that fluid-activated slippage in faults can quickly outpace the spread of fluid underground.

“One important constraint in developing reliable numerical models of seismic hazard is the lack of observations of fault behavior in its natural habitat,” said Pathikrit Bhattacharya, a former post-doc in the department of civil and environmental engineering at Tufts University’s School of Engineering and lead author of the study. “These results demonstrate that, when available, such observations can provide remarkable insight into the mechanical behavior of faults and force us to rethink their hazard potential”. Bhattacharya is now assistant professor in the School of Earth, Ocean and Climate Sciences at the Indian Institute of Technology in Bhubaneswar, India.

The hazard posed by fluid-induced earthquakes is a matter of increasing public concern in the US. The man-made earthquake effect is considered responsible for making Oklahoma— a very active region of oil and gas exploration—the most productive seismic region in the country, including California. “It’s remarkable that today we have regions of man-made earthquake activity that surpass the level of activity in natural hot spots like southern California,” said Robert C. Viesca, associate professor of civil and environmental engineering at Tufts University’s School of Engineering, co-author of the study and Bhattacharya’s post-doc supervisor. “Our results provide validation for the suspected consequences of injecting fluid deep into the subsurface, and an important tool in assessing the migration and risk of induced earthquakes in future oil and gas exploration.”

Most earthquakes induced by fracking are too small -- 3.0 on the Richter scale -- to be a safety or damage concern. However, the practice of deep injection of the waste products from these explorations can affect deeper and larger faults that are under stress and susceptible to fluid induced slippage. Injection of wastewater into deep boreholes (greater than one kilometer) can cause earthquakes that are large enough to be felt and may cause damage.

According to the U.S. Geological Survey, the largest earthquake induced by fluid injection and documented in the scientific literature was a magnitude 5.8 earthquake in September 2016 in central Oklahoma. Four other earthquakes greater than 5.0 have occurred in Oklahoma as a result of fluid injection, and earthquakes of magnitude between 4.5 and 5.0 have been induced by fluid injection in Arkansas, Colorado, Kansas and Texas.

原始論文：Pathikrit Bhattacharya, Robert C. Viesca. Fluid-induced aseismic fault slip outpaces pore-fluid migration. Science, 2019; 364 (6439): 464 DOI: 10.1126/science.aaw7354

引用自：Tufts University. “Study suggests earthquakes are triggered well beyond fluid injection zones.”