考虑关断时间的地面瞬变电磁三维带地形反演

起伏地形和关断时间对地面瞬变电磁响应影响严重,这给传统基于水平地表模型和理论阶跃波形的瞬变电磁数据解释技术带来很大困难.为此,本文开展考虑起伏地形和关断时间的地面瞬变电磁三维反演算法研究.正演采用基于非结构网格和后退欧拉隐式时间离散格式的时间域有限元算法,快速模拟起伏地表模型瞬变电磁响应.反演采用L-BFGS算法,减少每次反演迭代的计算量.针对发射线圈随地表起伏变化的特点,利用基于偶极子离散的场源处理技术模拟发射源的实际形状,采用瞬时电流脉冲技术实现考虑关断时间的地面瞬变电磁三维正演模拟.我们首先将本文开发的三维反演算法应用于理论模型的反演计算中,检验本文算法的可靠性,并分析地形和关断时间对反演结果的影响特征.在此基础上,进一步将本文算法应用于实测数据反演,验证本文算法的实用性.     

考虑发射波形的瞬变电磁三维模型降阶快速正演算法

瞬变电磁响应受到发射波形的影响,正演计算中需考虑发射波形.基于位移逆Krylov子空间投影的模型降阶算法能够精确模拟三维全波形瞬变电磁正演响应,但该算法计算精确的on-time响应非常耗时.目前实际工作中多数情况下都是对off-time时间段的数据进行处理解释,因此可以将正演问题简化为精确模拟考虑发射波形的off-ti...     

多道瞬变电磁法共中心点道集数据联合反演

多道瞬变电磁法是目前地面资源勘查领域的研究热点,开展该方法观测数据的一维反演对完成装备研发及技术推广具有积极的推动作用.针对多道瞬变电磁法特殊的源波形和数据采集方式,本文对多道瞬变电磁数据的响应提取和一维反演进行了研究.首先,采用反卷积方法从伪随机二进制序列全波形响应数据中提取大地脉冲响应.为了避免脉冲响应中空气波的干扰,提出采用反向积分算法,把脉冲响应转换成下降沿阶跃响应,并将其整理成共中点道集.最后,为充分利用各个偏移距下的响应数据对不同深度范围目标体的分辨能力,采用基于OCCAM算法的联合反演方法对共中心点道集数据进行联合反演.结果表明在不同目标体深度以及有噪声干扰的情况下,共中心点道集数据联合反演计算均可获得较好的结果.     

多通道瞬变电磁法2D有限元模拟

多通道瞬变电磁法的反演工作是在大地脉冲响应的基础上进行的,因此本文用2D有限元正演来求取大地脉冲响应,实现多道瞬变电磁法的正演模拟研究.研究工作首先在有源大地电磁法2D有限元模拟的基础上进行,原有2D有限元法的频带相对于多通道瞬变电磁法的频带是窄频的,为此采用校正法将其扩展到所需的宽度,实现了频率域宽频带有源电磁勘探方法的正演模拟;然后通过频时变换变换到时间域,得到瞬变电磁法的阶跃响应;最后通过求取阶跃响应的时间导数,得到大地脉冲响应.针对多通道瞬变电磁方法的装置特点,用有一定埋深的顺层成矿模型分析了大地脉冲响应特征,发现对有一定埋深的顺层产出的矿脉模型来说,频率域电磁场结果以及时间域大地脉冲响应结果均需移除背景场才能突出异常体的存在,然而,在实际工作中,未知的背景场限制了其实用性.为此,本文利用多通道瞬变电磁方法数据量大的特点,提出用不同偏移距的等时曲线与等时断面来展示脉冲响应模拟结果,事实表明这两种展示方式可以清晰地分辨出矿体引起的异常,且能准确定位矿体的中心在地面的投影位置,从而说明多通道瞬变电磁方法相对于频率域有源电磁方法,对有一定埋深的顺层矿床有更高的分辨能力.     

多通道瞬变电磁有限差分正演模拟

为适应多通道瞬变电磁法的特点,文章对Wang和Hohmann的经典时域算法进行了四点改进:第一,通过解一个三维直流电法正演问题获取准确的初始值,既避免了光滑波形近似带来的误差,又可适应浅表电阻率不均匀的模型;第二,利用阶跃响应和脉冲响应的导数关系对二者进行交错递推,可同时获得阶跃响应和脉冲响应而不增加计算量;第三,引入了Kirchhoff积分延拓公式,代替传统的基于静磁场假设的地空边界条件,提高了地空边界条件的精度.第四,通过褶积合成最终响应,可以适应任意发射电流波形.文中利用改进后的三维时域差分格式分别对均匀半空间模型、含油气薄层的层状模型和含油气薄层的盐丘模型进行了正演计算,并通过与伪随机码褶积合成了带波形的响应.计算结果表明:在合适的参数条件下,该差分格式是稳定的,且模拟结果是精确可靠的.含油气薄层的模型计算结果证明多通道瞬变电磁法对油气薄层有较好的勘探能力.     

多通道瞬变电磁法资料处理方法技术综述

多通道瞬变电磁法是近年来国际发展的油气资源探测新方法,为了实现这种方法由油气资源领域向矿产资源勘查领域的拓展,中国科学院地质与地球物理研究所牵头对其装备与数据处理解释系统开展技术攻关.针对上述自主研发的多通道瞬变电磁装置,基于其数据采集、数据处理与解释等技术要求,本文对与之匹配开发的多通道瞬变电法磁数据处理解释系统(包含预处理、大地响应正演模拟、二维和三维数据偏移成像等模块)的性能进行分析,结果表明:经过对多通道瞬变电磁法数据处理方法的技术攻关,形成了包含数据处理、正演模拟、二三维偏移成像、解释与可视化等模块的多通道瞬变电磁数据处理解释系统,为多通道大功率电法勘探系统的应用提供了良好的支撑作用.     

利用多通道瞬变电磁法识别深部矿体——以内蒙兴安盟铅锌银矿为例

多通道瞬变电磁法由于采用伪随机编码发射、最佳匹配相干解码技术,所以相比于传统时间域电磁方法勘探深度大且抗干扰能力强,是目前地面资源勘查领域的研究热点.针对内蒙兴安盟某铅锌银矿埋藏深度大等特点,首先根据矿区资料建立一维以及二维模型,数值计算脉冲响应并分析其特征,说明多通道瞬变电磁对于大埋深低阻异常体有良好的反映,随后通过多通道瞬变电磁技术对矿区进行实例勘查,数据处理过程中利用了反卷积的方法提取大地脉冲响应,根据轴向大地脉冲响应的峰值时刻计算视电阻率,通过对视电阻率进行全波形反演得到解释剖面,与测区钻孔资料对比分析证明,多通道瞬变电磁法是反映深部低阻异常体的有效手段.     

多道瞬变电磁法(MTEM)国外研究进展

多道瞬变电磁法(Multi-transient electromagnetic method,MTEM)最早由英国爱丁堡大学学者提出,该方法不仅融合了地震勘探中的数据采集和处理解释方法,而且把电磁法勘探中的几何测量和感应测量结合在一起,它是目前石油和矿产资源勘查领域具有发展潜力的电磁勘探方法.本文首先对MTEM方法的提出和发展过程进行了回顾,然后分析了其采用大功率接地源形式、伪随机码发射源、多道观测、类地震资料处理等特点等.最后介绍了采用MTEM在陆地、海洋等勘探环境下进行的探测和监测实际勘探例子.该方法成功的应用,为推进我国多道瞬变电磁深部探测系统研发起到了积极的作用.     

全空间条件下矿井瞬变电磁法粒子群优化反演研究

煤矿井下矿井瞬变电磁法(MTEM)探测中,电磁场呈全空间分布,全空间瞬变电磁反演是复杂的非线性问题,目前反演计算中全空间响应主要由半空间响应乘以全空间响应系数来得到,导致反演结果中顶板和底板异常(或前方和后方异常)叠加在一起难以分离,造成分辨率下降.论文提出采用粒子群优化算法(PSO)进行全空间MTEM反演,通过理论分析,在常规的粒子群算法基础上提出了一种新的进化公式改进策略,提高了粒子群算法的寻优能力.基于全空间瞬变电磁场理论,编写了粒子群算法反演程序,进行全空间条件下五层含巷道的复杂模型的反演计算.结合某矿井巷道顶板、底板岩层及断层含水性的探测实例,对实测数据进行反演计算和解释,探测结果得到钻探证实.研究表明,改进的粒子群优化算法对理论模型和实际资料的反演拟合程度较高,实现了矿井顶板、底板视电阻率异常的分离,提高了全空间瞬变电磁勘探资料的解释精度和分辨率.     

基于瞬时电流脉冲的三维时间域航空电磁全波形正演模拟

传统时间域航空电磁全波形正演模拟主要采用间接法(褶积算法)和直接法(时域有限差分方法等),然而褶积算法需要获得精确的电流二阶导数,这给发射电流数据采集工作带来极大挑战;时域有限差分方法受到网格和时间步长的严格限制,缺乏灵活性.为解决这些问题,本文采用时域有限元方法,通过直接改变每个时间道上的瞬时电流强度模拟任意发射波形的电磁响应.由于无需计算电流二阶导数,大大提高了正演结果的精度.利用基于非结构四面体网格的矢量有限元方法和后推欧拉技术对时间域电场扩散方程进行空间和时间离散,实现三维航空电磁时间域全波形的直接正演模拟.由此不仅可以模拟复杂的地电结构,而且基于后推欧拉法的无条件稳定性,可以更加灵活地选取时间步长,提高计算效率.通过与1D数值模拟结果进行对比验证了该方法的准确性.本文对三维柱状体模型上HELITEM MULTIPULSE和VTEM系统实际发射波形电磁响应进行模拟,并与褶积算法的结果进行比较,验证了本文算法模拟实际发射波形电磁响应的优越性.对复杂三维地质体模型上不同发射波形电磁响应进行模拟,验证了时间域有限元算法可有效处理复杂地下地质结构.     

The marine controlled source electromagnetic response of a steel borehole casing: applications for the NEPTUNE Canada gas hydrate observatory

Gas hydrates are a potential energy resource, a possible factor in climate change and an exploration geohazard. The University of Toronto has deployed a permanent seafloor time‐domain controlled source electromagnetic (CSEM) system offshore Vancouver Island, within the framework of the NEPTUNE Canada underwater cabled observatory. Hydrates are known to be present in the area and due to their electrically resistive nature can be monitored by 5 permanent electric field receivers. However, two cased boreholes may be drilled near the CSEM site in the near future. To understand any potential distortions of the electric fields due to the metal, we model the marine electromagnetic response of a conductive steel borehole casing. First, we consider the commonly used canonical model consisting of a 100 Ωm, 100 m thick resistive hydrocarbon layer embedded at a depth of 1000 m in a 1 Ωm conductive host medium, with the addition of a typical steel production casing extending from the seafloor to the resistive zone. Results show that in both the frequency and time domains the distortion produced by the casing occurs at smaller transmitter‐receiver offsets than the offsets required to detect the resistive layer. Second, we consider the experimentally determined model of the offshore Vancouver Island hydrate zone, consisting of a 5.5 Ωm, 36 m thick hydrate layer overlying a 0.7 Ωm sedimentary half‐space, with the addition of two borehole casings extending 300 m into the seafloor. In this case, results show that the distortion produced by casings located within a 100 m safety zone of the CSEM system will be measured at 4 of the 5 receivers. We conclude that the boreholes must be positioned at least 200 m away from the CSEM array so as to minimize the effects of the casings.     

伪随机编码源电磁响应的精细辨识

与传统阶跃源激励方式相比,采用m序列伪随机编码对发射源波形进行编码,提高了电磁探测的深度和分辨能力.然而受这种编码源发射波形自相关旁瓣效应的影响,使得对大地冲激响应的精细辨识效果受到一定限制.为了解决这一问题,在以往相关辨识方法研究的基础上,进一步考虑发射自相关旁瓣的影响,首先提出一种由收发互相关中高精度提取大地冲激响应的数学方法;然后通过数值模拟给出了m序列编码源大地冲激响应的精细辨识结果;同时对以m序列为发射波形的勘探系统相关参数选择进行了分析;最后利用本文提出的方法对野外实测数据进行辨识处理,通过与其他EM方法结果进行对比,证明了本文提出方法的可靠性.     

Grounded-source TEM modelling of some deep-seated 3D resistivity structures

Long-offset transient electromagnetics (LOTEM) is now regarded as a suitable electrical method for deep exploration along with magnetotellurics (MT). In this method, the vertical magnetic-field impulse response and, occasionally, the horizontal electric-field step response of a grounded-wire source on the surface of the earth are measured. Here, these two responses are computed for 3D models of three deep resistivity structures of interest in hydrocarbon exploration: (i) a faulted graben in a resistive basement rock at a depth of 4 km beneath a conductive overburden; (ii) a facies change in a resistive layer buried at a depth of 2 km in the conductive overburden above a resistive basement; and (iii) an anticlinal uplift of a resistive layer at a depth of 1 km in the conductive overburden above a resistive basement. The results show that the sensitivity of the electric-field response to model perturbation is generally greater than that of the magnetic-voltage response. Further, the electric-field sensitivity is confined to early and intermediate times while that of the magnetic-voltage response is confined to intermediate and late times. Hence it is recommended that both electric and magnetic recordings are made in a LOTEM survey so that the final results can be presented as apparent-resistivity curves derived from the two responses jointly as well as separately.     

A scaled experiment for the verification of the SeaBed Logging method

SeaBed Logging (SBL) is an application of the marine controlled source electromagnetic (CSEM) method that is used to directly detect and characterize possible hydrocarbon-bearing prospects. Although the CSEM method has been used by academia for more than three decades, the application as a direct hydrocarbon indicator was first introduced about five years ago. The central idea of SBL is the guiding of electromagnetic energy in thin resistive layers within conductive sediments. Even if it has been well known for a long time that electromagnetic signals can propagate from a conductive region to another via resistive regions such as air or resistive parts of the lithosphere, the application to hydrocarbon exploration has not been developed until recently. This might be due to the uncertainty of getting any significant response from thin resistive layers such as hydrocarbon reservoirs since electromagnetic energy is highly attenuated in conductive sediments. Thus, during the early development phase of the SBL technique, a scaled laboratory experiment was performed to validate if a thin resistive layer (e.g. hydrocarbons) buried within conductive media (e.g. sediments) could be remotely detected by using electric dipoles as sources and receivers. Data from this experiment were compared to a forward modelling code for layered media, and the comparison showed good agreement between experimental and theoretical results. This suggested that thin resistive layers buried in conductive media are detectable due to the guiding of the electromagnetic field within the resistor. The successful results were vital for realizing the application of marine CSEM as a hydrocarbon exploration technique. We here present the results of the first scaled SBL experiment.     

时间域航空电磁系统瞬变全时响应正演模拟

近年来很多专家学者致力于时间域航空系统正反演的研究.本文针对一维均匀层状介质和三维模型进行正演.不仅计算垂直方向上的电磁响应,还计算了同线方向上的电磁响应,为航空电磁多分量观测提供理论依据.通过比较航空电磁系统的脉冲响应和阶跃响应特征,发现脉冲响应在早期时间存在奇异性,造成计算不稳定.然而,阶跃响应在早期时间没有奇异性,因而利用系统的阶跃响应可得到一种计算时间域航空电磁系统全时响应的稳定算法.该算法具有较高的精度,并很好地保持了磁场强度B和磁感应dB/dt关系的一致性.该算法推广到三维地质体的时间域正演模拟亦取得很好的效果.     

A numerical comparison of time and frequency‐domain marine electromagnetic methods for hydrocarbon exploration in shallow water

In shallow water the frequency domain controlled source electromagnetic method is subject to airwave saturation that strongly limits the sensitivity to resistive hydrocarbon targets at depth. It has been suggested that time‐domain CSEM may offer an improved sensitivity and resolution of these deep targets in the presence of the airwave. In order to examine and test these claims, this work presents a side‐by‐side investigation of both methods with a main focus on practical considerations, and how these effect the resolution of a hydrocarbon reservoir. Synthetic noisy data for both time‐domain and frequency domain methods are simulated using a realistic frequency dependent noise model and frequency dependent scaling for representative source waveforms. The synthetic data studied here include the frequency domain response from a compact broadband waveform, the time‐domain step‐response from a low‐frequency square wave and the time‐domain impulse response obtained from pseudo‐random binary sequences. These data are used in a systematic resolution study of each method as a function of water‐depth, relative noise and stacking length. The results indicate that the broadband frequency domain data have the best resolution for a given stacking time, whereas the time‐domain data require prohibitively longer stacking times to achieve similar resolution.