频谱差异烃类检测新方法在东胜气田的应用
A new method of spectrum difference hydrocarbon detection and its application to the Dongsheng gas field
责任编辑: 叶佩
收稿日期: 2020-03-30 修回日期: 2020-05-28 网络出版日期: 2020-08-20
基金资助: |
|
Received: 2020-03-30 Revised: 2020-05-28 Online: 2020-08-20
作者简介 About authors
王东辉(1982-),男,2010年毕业于长江大学矿产普查与勘探专业,主要从事地质、地震解释与管理工作。Email:
查明含烃区与不含烃区的频谱差异是利用频谱属性开展地震烃类检测工作的重要前提。为了直观地确定出含烃区与不含烃区的频谱差异,提出基于频谱差异的烃类检测方法,结合实际油气试采数据和研究需要,将已知井分为含烃井与不含烃井,通过比较分类对象射孔段的瞬时振幅谱(点谱)来找出含烃相关的频谱差异,并利用多种频谱属性交会的方式得到地震含烃区。本文通过目的层盒1段下伏煤层有无的正演模型确定了最大波谷振幅属性为预测储层信息的有利属性,但更深入分析认为频谱差异属性比常规的最大波谷属性更能表征含烃与不含烃的差异。通过设置属性门槛值的方式来量化含烃区与不含烃区在频谱属性上的差异,并将10 Hz单频振幅、25 Hz积分振幅、低频衰减梯度、高频衰减梯度等4种含烃门槛值内的频谱属性进行交会,继而确定出潜在的含烃区,通过水平井验证检测结果,表明该方法具有一定的有效性。
关键词:
The crucial precondition of using spectral attributes to conduct hydrocarbon detection is to investigate the spectral differences between gas-bearing and none-gas bearing area.In order to determine the spectrum difference between hydrocarbon bearing and none-hydrocarbon bearing areas intuitively and incorporate the constraints of oil and gas production testing data,this paper proposes a hydrocarbon detection technology based on spectrum difference,which divides known wells into hydrocarbon-bearing wells and non-hydrocarbon bearing wells according to the actual oil and gas production testing data and research needs.By comparing the instantaneous amplitude spectral of pay zones of classified objects,researchers can recognize spectral differences related to hydrocarbons and use the intersection of multiple spectrum attributes to characterize the differences so as to obtain the seismic hydrocarbon-bearing areas that could be identified.The maximum trough amplitude was determined as the relative advantage attribute to recognize the reservoir information of the H1 member according to the forward modeling contrast between presence and absence of underlying coal seams.More in-depth analysis shows that the spectrum difference can better forecast hydrocarbon-bearing and hydrocarbon-absent zones than the maximum trough amplitude.The differences of spectral attributes between hydrocarbon-bearing areas and non-hydrocarbon bearing areas was quantified by setting attributed threshold value,and the spectral attributes in individual thresholds of the four spectral attributes were crossplotted to dentify the potential hydrocarbon bearing areas.The hydrocarbon dictation results are supported by horizontal wells,which indicates the method is reliable.
Keywords:
本文引用格式
王东辉, 吴晓川.
WANG Dong-Hui, WU Xiao-Chuan.
0 引言
地震烃类检测是油气勘探人员长期关注的研究领域。从早期提出的亮点、暗点和平点技术等直接烃类指示技术到AVO技术,再到谱分解技术[13],这一系列技术的相继提出反映了地震烃类响应的多样性和复杂性以及研究人员思路的开阔性。谱分解衍生的频谱属性作为地震属性的一种,它被定义为从地震子波振幅谱或功率谱所衍生出来的属性,任何可用谱函数表征的参数均可称之为频谱属性[4],例如振幅谱的峰值频率、峰值振幅、积分振幅属性等。频谱属性在1970年提出之时被用于上地幔的研究[5],90年代以后被应用于地震油气勘探中[6-10]。频谱属性后期被广泛用于烃类检测,依据含烃前后频谱形态的变化(高频衰减低频增强、主频降低)发展出多种烃类检测方法:面积差值法[11]、衰减梯度法[12]、谱比法[13]、频谱衰减法[14]、流体活动因子法等[15]。应用这些方法进行烃类检测虽然取得了一定的效果,但它们倚重于地震资料频谱属性的计算,在存在一定数量的探井且勘探程度相对较高的油气田内,直接运用上述方法就显得与实际的油气试采资料结合得不够密切。此外,在一些报道中地层含烃后的变化与常规的认识出入较大,如低频衰减高频增强、主频升高等[9,16-17],直接使用上述方法也存在一定的不确定性。
通过对鄂尔多斯盆地东胜气田盒1段有无下伏煤层的正演模型确定了最大波谷振幅属性为表征盒1段储层信息的相对有利属性。根据试气结果将17口探井分为含烃井(11口)与不含烃井(6口),盒1射孔段频谱分析表明含烃井10~25 Hz内的单频振幅和积分振幅普遍比不含烃井的单频振幅和积分振幅要高出许多。提取并优选出10 Hz单频振幅和25 Hz积分振幅来表征盒1段含烃区与不含烃区的频谱差异,此外,低频衰减梯度与高频衰减梯度属性亦能较好地表征这种差异。对比发现优选出的频谱属性比受煤系反射影响较弱的最大波谷振幅属性更能体现含烃与不含烃的差异,因此舍弃了常规属性的使用。通过设置属性门槛值的方式来量化含烃区与不含烃区在频谱属性上的差异,并将上述4种含烃门槛值内的频谱属性进行交会,继而确定出潜在的含烃区。利用水平井的钻遇情况来验证检测结果,表明该方法具有一定的有效性。
1 频谱差异烃类检测方法背景
面积差值法、衰减梯度法、谱比法和频谱衰减法、流体活动因子法的实质都是在探寻含烃频谱和不含烃频谱的差异(图1),它们的提出主要依赖于地层含烃源后高频快衰减、主频降低、低频增加、频带宽度变窄等[20]。然而,地层含烃后产生的频谱差异并不总是完全符合这些规律。在四川盆地须家河组致密气藏的地震烃类检测过程中,有学者发现一些气井射孔段的峰值频率和峰值振幅往往比非气井的要大[16]。当地震波穿透含气层段地层后,主频往高频方向移动,在区分产气井和不产气井方面频带宽度属性也未能起到效果[16]。邓继新等通过楔状模型顶底反射波的谱比论述了砂岩储层含气后主频的移动方向,就厚层砂岩处于低速围岩背景而言,地震波途经含气砂岩比途经含水砂岩产生更明显的降频效应;若砂岩与其上覆和下伏地层在垂向上组成反射系数增大的结构时,砂岩含气后主频反而会升高[21]。此外,对于厚度小于半波长的储层而言,地震波传播时间短而不能造成强烈的吸收衰减,不能形成明显的低频阴影,甚至会因气层顶底的干涉效应而产生高频增强的反射,这已被实际的油藏地震资料所揭示并证实[9,17]。因此,地层含烃后造成的频谱差异具有多样性,如何识别含烃区与不含烃区的频谱差异是烃类检测的关键环节。
图1
图1
面积差值法、衰减梯度法、频谱衰减法、流体活动因子法的图示及主要依据
Fig.1
The basic for the area difference method,attenuated gradient method,spectral attenuated method and fluid flow method
2 频谱差异烃类检测方法的思路
应用频谱属性进行地震烃类检测更多的是基于高频快衰减低频慢衰减的原理,但真实的情况比这更为复杂,主频既能往低频方向移动,也能往高频方向移动。储层含烃后无论表现为常规的衰减形式还是非常规的衰减形式,都表明含气区与不含气区的频谱差异是客观真实存在的,因而运用频谱属性进行烃类检测就是要探寻并表征这种差异。所以,对于一个存在油气井、水井和干井的勘探开发区块,研究人员可采取分类统计的方法划分出含烃井与不含烃井(井数量多的条件下,可根据油气产量进行更细的划分),拾取它们射孔段上的瞬时振幅谱(点谱),以实际观察到的频谱差异作为烃类检测的依据,这样既保证了地层含烃后形成的特殊频谱差异,同时也融入了油气试采数据的约束,使得烃类检测的实际意义更为突出。
3 频谱差异烃类检测方法应用实例
图2
图2
东胜气田三维区位置(a)和三维区内井位分布及盒1段气水柱状(b)图
Fig.2
The regional map of 3D seismic area in Dongsheng gas field(a) and the distribution of well locations and gas-water column diagram of the H1 member in 3D survey(b)
3.1 盒1段有效反射信息及频谱敏感属性分析
图3
图3
盒1段过J77、J120、J123井连井剖面及盒1段射孔位置
Fig.3
The well section profile of the H1 member,including well J77,well J120 and well J123, showing the location of the payzones of the H1 member
图4
图4
过J77、J120、J121井地震剖面与层位标定
Fig.4
The seismic section crossing well J77,well J120 and J121,marked with horizons
图5
图5
无煤层(a)和有煤层(b)情况下的正演模拟
Fig.5
Forward model with coal seams (a) and without coal seams (b)
图6
图6
盒1段均方根振幅(a)和T9d波峰之上波谷的最大绝对值振幅(b)
Fig.6
The RMS of the H1 member(a) and the maximum absolute value amplitude of the trough above the T9d peak(b)
图7
图7
盒1段井点均方根振幅(a)和T9d波峰之上最大波谷振幅(b)统计直方图
Fig.7
Histogram of the RMS attribute at the well location corresponding to the H1 member(a) and histogram of the maximum absolute value amplitude at the well location corresponding to the trough above the T9d peak(b)
相反,含气井位处的最大波谷属性值比不含气井位处的最大波谷属性值要高出许多,二者的重叠范围较小(图7b),数值上区分明显。综合上述的分析可知,盒1段的最大波谷属性比盒1段的均方根振幅属性更适用于含气检测。
3.2 盒1段频谱差异相关属性的优选
根据井上盒1段储层的空间分布位置和时间域与T9d波组的关系,分别提取含气探井与不含气探井射孔段上的瞬时振幅谱和积分振幅谱(采用广义S变换)。从图8a可以看出,同不含气井相比,含气井的频带范围更宽,振幅值也相对较大。峰值振幅方面,除不含气井J106外,含气井普遍高于非含气井。同样地,在10~25 Hz区间段,含气井的振幅更强,超过25 Hz后,含气井与不含气井的频谱曲线混杂交织在一块,差异不明显(图8b)。从图8b可以看出,10 Hz以下,含气井与不含气井的积分振幅差异不明显。在10~25 Hz区间内,这两者之间的差异逐渐显现,含气井的积分振幅显著增大,不含气井中仅有J56和J104的积分振幅值高于含气井的积分振幅值,其余不含气井的积分振幅值均小于含气井的积分振幅值。超过25 Hz后,含气井J94与J51就同不含气井的积分振幅属性值曲线发生交叉与重叠,甚至小于不含气井的积分振幅值(图8b)。
图8
图8
含气井与不含气井盒1射孔段的瞬时振幅谱图(a)和积分振幅谱图(b)
Fig.8
Instantaneous amplitude spectrum (a) and integrated amplitude spectrum (b) of the H1 member corresponding to gas-bearing wells and non-gas-bearing wells
图9
图9
盒1段含气井与不含气井井位处10 Hz(a)、15 Hz(b)、20 Hz(c)、25 Hz(d)单频振幅属性值直方图
Fig.9
Histogram of the isolated amplitude at the well locations of the H1 member, corresponding to 10 Hz (a), 15 Hz (b), 20 Hz (c), 25 Hz (d), respectively
图10
图10
盒1段含气井与不含气井井位处10 Hz(a)、15 Hz(b)、20 Hz(c)、25 Hz(d)积分振幅属性直方图
Fig.10
Histogram of the integrated amplitude at the well locations of the H1 member,corresponding to 10 Hz (a),15 Hz (b),20 Hz (c),25 Hz (d),respectively
图11
图11
盒1段含气井与不含气井井位处低频衰减梯度直方图(a)与高频衰减梯度属性直方图(b)
Fig.11
Histogram of the low frequency attenuated gradient(a) and high frequency attenuated gradient(b) at the well locations of the H1 member
通过对5、10、15、25 Hz的单频振幅属性和积分振幅属性进行优选,优选的依据是在含气井与不含气井的属性值重叠区间内,不含气井的个数要越少越好。含气井的属性值要具有向高值集中的趋势,以保证含气井与不含气井的属性值差异最大化。从预测期望上来看,希望含气井属性值和产气量都要大。因此,本次优选出了10 Hz单频振幅和25 Hz的积分振幅作为表征频谱差异的最佳属性。
3.3 频谱属性与最大波谷振幅属性的对比
3.4 盒1段频谱差异区的平面表征
按照上述优选出来的属性,合理设立属性门槛值,以此来量化含气井与不含气井的差异。针对盒一段的频谱差异进行门槛值的确定和最终的交会分析。井点处10 Hz的振幅属性统计直方图上可见J20井比其他不含气井的振幅值要小很多,且其产气量仅有0.145×103 m3·d-1,远小于其他含气井的产量(图2b)。因此,在确定10 Hz振幅属性门槛值时,若将门槛值的下限设定在30以下包括J20井,则包含了许多不含气井,故将J20井视为不含气井,以便减少门槛值范围内不含气井的个数。因而将10 Hz振幅中区分含气井与不含气井的门槛值区间设定在60~160(图9a),此门槛值范围内仅有1口不含气井(J122)。同样地,25 Hz积分振幅中区分含气井与不含气井的门槛值区间设定在4.5×107~3.7×108(图10d),该范围内存在3口不含气井(J121、J16和J106);低频梯度属性中区分含气井与不含气井的门槛值区间设定在1.0×106~1.7×107(图11a),在这范围内仅存在1口不含气井(J121);高频梯度属性中区分含气井与不含气井的门槛值区间设定在3.0×106~1.65×107(图11b),该区间内存在2口不含气井(J121和J106)。把这些介于门槛值区间内的频谱属性进行交会(图12),就可得到门槛值区间内的交会属性分布图(图13)。图13 为下石盒子组盒1段气水柱子与频谱差异交会分析平面叠合图,白色区为不含气区,彩色区为门槛值范围内的地震含烃区域。从图13分析,J77P1H井不在本次交会出的区域中,J77P2H、J77P3H、J77P4H、J77P5H井均落在了交会区域,这表明,该研究方法和思路具有一定的可靠性。
图12
图12
门槛值内25 Hz积分振幅、低频衰减梯度、高频衰减梯度和10 Hz的单频振幅属性交会流程
Fig.12
The crossploting flow diagram of the attributes within the threshold value, including integrated amplitude of 25 Hz, the low frequency attenuated gradient attribute, the high frequency attenuated attributed and the 10 Hz isolated amplitude
图13
图13
盒1段气水柱子与频谱差异交会分析平面叠合图
Fig.13
The superimposed map of the gas-water column of the H1 member and spectral difference area,white area indicating the non-gas-bearing area while colorful area indicating the potential gas-bearing area revealed by seismic data
4 结论
直接使用衰减面积、衰减梯度、谱比属性等方法进行烃类检测具有一定的风险且缺乏实际油气试采数据的约束。在有油气试采数据的前提下,针对已知井的射孔段,分析频谱特征差异与含油气性的关系,建立含油气性识别预测模式。在井少的情况下可作为参考依据,井多的条件下则更具有地质统计意义。
1)鄂尔多斯盆地东胜气田下石盒子组盒1段受下伏煤系反射影响严重,盒1段地震烃类检测研究表明,在有无煤层正演模型和井点属性值的统计分析认为, 最大波谷振幅属性受煤系地层地震反射影响弱,比盒1段均方根振幅属性更有效,但不如优选出来的盒1段频谱属性更能反映含油气的特征。
2)鄂尔多斯盆地东胜气田下石盒子组盒1段区分含气井与不含气井的优势频谱属性为10 Hz单频振幅、25 Hz积分振幅、低频和高频衰减梯度属性。由这4种所选门槛值范围之内的频谱属性交会出该区盒一段地层潜在含气分布范围,其中,在5口水平井中有4口水平井落在了交会的含气区域,表明检测结果具有一定的可信度。
3)对储集层或目的层从采用频谱聚类的方法,如使用无监督频谱聚类,对比已知井的关系来检测油气,也可利用已知井点建立样本,开展有监督频谱聚类以检测烃类分布。
参考文献
Detection of gas in sandstone reservoirs using AVO analysis:A 3-D seismic case history using the Geostack technique
[J].DOI:10.1190/1.1443695 URL
Seismic bright spots in the Gulf of Oman
[J].DOI:10.1016/0012-821X(77)90143-1 URL
Application of spectral decomposition to detection of dispersion anomalies associated with gas saturation
[J].DOI:10.1190/1.2172314 URL
Dependence of theP-wave amplitude spectrum on focal depth
[J].DOI:10.1007/BF00875887 URL [本文引用: 1]
Interpretational applications of spectral decomposition in reservoir characterization
[J].DOI:10.1190/1.1438295 URL [本文引用: 1]
Narrow-band spectral analysis and thin-bed tuning
[J].DOI:10.1190/1.1487075 URL
Spectral attributes for attenuation analysis in a fractured carbonate reservoir
[J].DOI:10.1190/1.1518443 URL
Application of spectral decomposition to gas basins in Mexico
[J].DOI:10.1190/1.1634918 URL [本文引用: 2]
Instantaneous spectral analysis:Detection of low-frequency shadows associated with hydrocarbons
[J].DOI:10.1190/1.1559038 URL [本文引用: 1]
基于广义S变换的吸收衰减分析技术在油气识别中的应用
[J].
Application of absorption and attenuation analysis technique based on generalized Stransform for hydrocarbon identification
[J].
Energy absorption analysis:A case study
[C]//
利用地震频谱比进行河道砂体预测
[J].
Application of seismic ratio method to predict channel sandbody
[J].
地震波频谱衰减检测天然气技术应用研究
[J].由于地层含油气后对高频成分吸收增强、吸收系数增大,这为利用吸收系数预测油气及含油气范围提供了理论依据。实验表明,地层对地震波的吸收作用主要取决于岩石骨架的弹性性质、岩石的孔隙率及孔隙中含流体成分等,因此沿反射层位计算地震波的吸收性质及其横向变化可以用来预测岩性和含油气性。文中利用基于模型的参量法来估算地震波频谱,其优势在于即使在小时窗也可以提取到准确的地震波频谱。具体方法包括以下步骤:①利用目的层顶面的地震追踪层位拾取目的层地震相位;②沿目的层上、下各开一个等大的时窗分别计算信号谱和子波谱;③求取上、下时窗的子波对数谱的绝对值比值,得到目的层含气影响所造成的频率吸收衰减谱。文中分别计算了研究区目的层上、下两个时窗内的地震频率信号谱及其各自的子波同态(Cepstral)谱,还比较了在经过两时窗所夹持的目的层后子波同态谱在有效频段内的吸收程度,认为在高频段的衰减主要由地层中含有天然气所致。]]>
Study on application of seismic frequency spectrum attenuation to detect natural gas
[J].
Seismic low‐frequency effects from fluid‐saturated reservoir
[C]//
Seismic spectral decomposition and analysis based on Wigner-Ville distribution for sandstone reservoir characterization in West Sichuan depression
[J].DOI:10.1088/1742-2132/7/2/002 URL [本文引用: 3]
High-frequency anomalies in carbonate reservoir characterization using spectral decomposition
[J].
DOI:10.1190/1.3554383
URL
[本文引用: 2]
Low-frequency shadows have often been used as hydrocarbon indicators in the application of spectral decomposition. The reason behind the low-frequency anomaly has been explained as high-frequency energy attenuation caused by hydrocarbons. However, in our practice on carbonate reservoir characterization in two areas, Precaspian Basin and Central Tarim Basin, China, we encountered high-frequency anomalies, i.e., the isofrequency slices or sections at high frequencies exhibit anomalies associated with the good carbonate reservoir, particularly in the tight limestone background. We used the product of porosity and thickness as a parameter to measure the quality of the carbonate reservoir of each well and classified the 46 wells in our two studied areas into three types. Type I wells contain high-porosity thick reservoirs, type II wells contain reservoirs with moderate porosity and thickness, and type III wells contain only low-porosity thin reservoirs. The results were that 12 out of 13 type I wells exhibit high-frequency anomalies, and 30 out of 33 type II and type III wells do not exhibit high-frequency anomalies. We further validated the existence of this high-frequency anomaly by forward modeling analysis and fluid substitution experiments using the actual well-log curves measured in the carbonate reservoir. The results showed that in our two studied areas the high-frequency anomalies are more common than low-frequency shadows that can be observed only when the thickness of the reservoir is more than half of the wavelength or the reservoir rocks are extremely unconsolidated. Therefore, this high-frequency anomaly may be used as a more reliable indicator for a good carbonate reservoir than low-frequency shadows in real applications.
薄储层(体)含油气性识别的HFC技术及其应用研究
[J].
The harmonic frequencies characteristics(HFC) analysis for identifying and predicting hydrocarbon in thin reservoir strata(bodies)
[J].
塔河油田缝洞储集体油水识别的谐频特征分析技术应用研究
[J].
Application of HFC technique for hydrocarbon identification infracture-cave reservoirs of Lower-Ordoician carbonate in Tahe Oilfeild
[J].
基于时变子波谱模拟的面积差值法吸收衰减属性研究与应用
[C]//
The research and application of absorption attenuation attribute of area difference method based on time-varying wavelet spectrum simulation property
[C]//
基于正演模型的储层谱分解响应特征分析
[J].
Analysis on spectral decomposition response of reservoir based on forward modeling
[J].
利用频谱积分属性评价碳酸盐岩储层
[J].
Evaluation of carbonate reservoirs by frequency spectrum integration attributes
[J].
油气检测技术在三湖浅层生物气勘探中的应用
[J].
Hydrocarbon detection in biogas exploration in Sanhu Shallow
[J].
频谱分解技术在泥页岩储层含气性预测中的应用——以柴达木盆地鱼卡凹陷为例
[J].
Application of spectral decomposition technology to the shale reservoir prediction:A case study of Yuqia sag in Qaidam Basin
[J].
地震可检测性分辨率研究
[J].
Research on seismic detectable resolution
[J].
Thickness imaging for high-resolution stratigraphic interpretation by linear combination and color blending of multiple-frequency panels
[J].
Spectra crossplot
[C]//
Three-dimensional seismic analysis of Late Paleozoic coal-bearing series reflections in the Hangjinqi,North Ordos Basin,China
[J].
/
〈 |
|
〉 |
