A Study on the Acquisition Technology for Weak Seismic Signals from Deep Geothermal Reservoirs

Abstract

There are rich geothermal resources in China and they are widely distributed. After years of continuous exploration and production, most of the shallow geothermal resources have been explored and the current exploration mainly focuses on deep ones. Among a great many geophysical exploration methods, seismic survey is the most effective means of geothermal resource exploration and production, but its weak seismic reflection signal, low S/N ratio and poor imaging of thermal reservoirs due to the seismic geological conditions restrict the production and utilization of deep geothermal resources. It considers through the analysis of the geophysical characteristics of the thermal reservoir that the main causes for the weak seismic reflection signals from deep thermal reservoirs are (1) the shielding effect of the strong wave impedance interface between the thermal reservoir and the caprocks; (2) the small reflection coefficient inside the thermal reservoir; and (3) the serious absorption and attenuation of high-temperature fluids such as hydrothermal fluids and steam in the thermal reservoirs by seismic waves. Accordingly, a series of seismic data acquisition technologies are proposed based on the high-precision vibroseis low-frequency shooting, high-sensitivity geophones long-spread receiving and small bin size highfold technologies, and their application in the seismic acquisition of dry-hot rocks (HDR) in Gonghe Basin, Qinghai, shows a very good application effect.

1. Introduction

Geothermal resource, as a new type of clean and renewable energy, has attracted widespread attention from all over the world. According to the official data released by the Ministry of Land and Resources of the People’s Republic of China, the deep hydrothermal geothermal resource in China is equivalent to 1.25 trillion tons of standard coal and the deep dry-hot rock resource in China is equivalent to 856 trillion tons of standard coal, which are 260,000 times the current total annual energy consumption in China. Its exploration and production are therefore of great significance. Geothermal energy is divided into three categories according to its buried depth and temperature: shallow geothermal resources, hydrothermal geothermal resources and enhanced geothermal resources (HDR geothermal resources). Shallow geothermal resources refer to the thermal energy stored in the rock and soil masses, groundwater and surface water within a certain depth (generally less than 200 m) at the surface. They are mainly utilized with heat pumps. Hydrothermal geothermal resources refer to the geothermal energy stored in groundwater. They are produced and utilized through natural channels or artificial drilling. HDR geothermal resources refer to the thermal storage of rock masses free of fluid or only with a small amount of underground fluid inside buried at 3000 m at a temperature > 200 °C, which is of great economic production value. With the continuous advancement of the “Green development strategy” in China, geothermal resource exploration has turned to deep exploration for larger resource reserves.

The exploration and evaluation of geothermal resources began in the 1950s in China. Through years of exploration and evaluation, a batch of medium–high temperature geothermal fields has been proven for the country and the basic conditions and distribution of geothermal resources shallower than 2000 m in the whole of China have been grasped preliminarily, forming a geothermal comprehensive geophysical exploration technology dominated by non-seismic survey and supplemented by seismic survey. The basement undulations of geothermal field (uplifts and depressions) and the spatial distribution of fault structures can be determined using the gravity method; the location of hydrothermal alteration zones, the distribution and thickness of hidden igneous rock bodies and their relationship with fault zones can be determined using the magnetic method; the thermal anomaly can be delineated and the range and depth of the thermal storage body can be determined using the electrical method; and the magma chambers and the thermal storage location and scale in high-temperature geothermal fields can be determined using the magnetotelluric method [1,2,3]. However, these non-seismic methods were unable to meet the fine exploration and production demands of deep geothermal resources due to their low accuracies. Seismic method, instead, as a high-precision ultra-deep exploration method, can accurately determine the location, occurrence and thermal storage structure of faults. It makes up for the shortcomings of gravity and magnetoelectric exploration to a large extent, becoming the most effective exploration and production means of geothermal resources. Nevertheless, the interlayer reflection coefficient is low for deep sedimentary geothermal reservoirs, due to the long-term deposition, relatively long geological age, changeable stratum lithology, the superposition of different tectonic movements in multiple periods and the impact of the compaction and the complexity of the inner stratigraphic structures,. There is usually a strong wave impedance interface between the deep old strata and the overlying strata and such a strong wave impedance interface has a strong shielding effect on the seismic signals from the underlying thermal reservoir. There is a strong wave impedance interface, which has a strong shielding effect on seismic waves, between deep igneous rock thermal reservoir and the surrounding rocks, making it difficult for the energy to propagate into the dry-hot rock mass, resulting in weaker reflected wave energy from the HDR layer on the inner side of the contact surface and the low S/N ration. In addition, the high-temperature HDR mass is generally in a quasi-massive structure with poor bedding and layering, and the interlayer reflection coefficient is small, resulting in weak reflected wave energy and a low S/N ratio of the formation inside the HDR mass. Such two types of reservoirs are characterized by weak reflected wave energy, a low S/N ratio and poor lateral continuity in the seismic data. Acquiring weak reflected signals from thermal reservoirs is a challenge for seismic survey.

Wang Qinghua and Han Wengong and others [4,5] studied the manifestation types and characteristics of weak signals. The results showed that weak signals were identifiable when the S/N ratio > 2; they were difficult to be identified when the S/N ratio = 1; and they were unidentifiable when the S/N ratio = 0.5. Zhi Wang, Ke’en Li, Ruizhen Wang et al. [6,7,8,9] proposed a long-spread wide-azimuth reflection seismic survey method for weak reflection signals shielded by high-velocity layers. Deping She [10,11,12,13,14,15] believed that low-frequency signals could be used to weaken the influence of the basalt high-velocity shielding layer and improve the deep imaging quality under the shielding layer significantly. Yanguang Wang, Wujin Chen et al. [16,17] proposed a high-precision high-density 3D seismic survey technology to solve the imaging problem of weak signals from deep volcanic rocks. Luzi Zhao, Ligui Xu et al. [18,19,20] believed that low-frequency broadband shooting, wide -line long-spread high-fold geometry recording and wide-line broadband high-precision 3D seismic survey technologies solved the deep—ultra-deep survey problem effectively and achieved good geological results. In this paper, aiming at the problems of weak reflection signals from deep thermal reservoirs, a series of seismic data acquisition techniques based on high-precision vibroseis low-frequency shooting, high-sensitivity geophone long-spread receiving and high-density acquisition are proposed and their application is shown taking the HDR survey in Gonghe Basin in Qinghai as an example, hoping to provide a reference for the acquisition of weak signals from deep thermal reservoirs.

2. Geophysical Characteristics and Weak Signal Acquisition Technologies

2.1. Geophysical Characteristics of Deep Thermal Reservoirs

In seismic survey, the intensity of reflected signals is mainly measured with the amplitude (waveform) of effectively reflected waves under a certain noise background. The so-called weak signal refers to the effective signal with such a small amplitude value under a certain noise background that it is difficult to be distinguished visually. The complex geological conditions, such as the type, scale, thickness, stratum and structural development, and the near-surface seismic geological conditions of geological targets are the main factors that cause the weak seismic reflection signals. There is a strong wave impedance interface, which has a strong shielding effect on seismic waves, between the deep igneous thermal reservoir and the surrounding rocks, making it difficult for energy to propagate into the HDR mass. Moreover, high-temperature HDR masses are generally in a quasi-massive structure, which is poorly layered. Disordered interfaces and their extension scale are small, the interlayer reflection coefficient is low. As shown in the composite logs in Figure 1, the velocity and wave impedance at the top and bottom of the granite interface (Expressed as Tg) change abruptly, corresponding to the strong reflection in the seismic record; the reflection coefficient gets smaller, the signals are weak and the reflection is disordered below Tg in the seismic record.

Sedimentary geothermal reservoirs are the result of a long geological age. The change in lithology of their formations and the superposition of multiple periods of different tectonic movements make the reservoir pattern very complicated: there is often a strong wave impedance interface between the old and the new strata and it has a very strong shielding effect on seismic waves; under the influence of compaction, the velocity difference between old strata and the interlayer reflection coefficient are small; under the influence of multistage tectonic movements, old strata are usually in complex structures with violent fluctuations.

Figure 2 shows the composite logs of a sedimentary hydrothermal well and the seismic section over the well. It can be seen from the figure that the formations above and below Tg in the red circle are similar to the above features in velocity, wave impedance and section seismic wave characteristics. Both types of thermal reservoirs have the problems of shielding of strong reflection interface and small internal reflection coefficient, both of which are reflected as weak reflected wave energy, low S/N ratio and poor lateral continuity in seismic data. In addition, the fractures occurring inside the thermal reservoirs further reduce the S/N ratio of data. Therefore, the task of seismic data acquisition is to adopt reasonable methods to overcome the shielding effect of the strong wave impedance interface, increase the S/N ratio of the weak reflection data inside the thermal reservoirs and improve the imaging effect of thermal reservoirs.

2.2. Weak Seismic Signal Acquisition Technologies

Weak seismic signal acquisition methods contain: (1) shooting: To increase the down transmission energy of seismic waves by increasing the charge size of explosive sources or using large-tonnage vibrators for low-frequency shooting; (2) receiving: To apply high-sensitivity geophones and long-spread receiving to improve the receiving of weak signals; and (3) geometry: Small bin (line) size is applied for high-density acquisition to reduce the loss of weak signals through sufficient, uniform and symmetrical sampling of the wave field.

2.2.1. The Low-Frequency Vibroseis Shooting Technology

The absorption and attenuation of seismic waves by the formation follow the law of G(f,t) = e^−π·ft/Q, where f refers to the frequency of seismic waves; t refers to the propagation time; and Q refers to the formation quality factor. It reflects that the absorption and attenuation of formation to seismic waves are related to frequency and they get stronger exponentially with the increase in frequency. Meanwhile, it also indicates that low-frequency signals have a strong anti-absorption ability, so they can penetrate the high-velocity shielding more easily to improve the deep imaging quality. Figure 3 shows the shape and amplitude of the Ricker wavelets of the dominant frequency 25 Hz and 50 Hz, respectively, that propagate through the same path in the same propagation time. The absorption attenuation of 50 Hz is significantly stronger than that of 25 Hz. Therefore, it is feasible to reduce the shielding effect of the strong wave impedance interface and improve the quality of the raw data using low-frequency shooting technology.

Based on the characteristics of the thermal reservoir, a model with granite as the thermal reservoir is established (see Figure 4a). The granite is divided into two layers, their velocities are 4800 m/s and 5400 m/s, respectively, and the velocities of their overlying sand and mudstone are 2100 m/s and 2900 m/s, respectively. The wave equation forward simulation is performed with the wavelets of different dominant frequencies: 3 Hz, 15 Hz, and 30 Hz, respectively (see the simulation results in Figure 4b–d). Seeing in the forward simulated single-shot record, the energy of the signal from the granite formation is strongest when shooting with 3 Hz, followed by 15 Hz and is weakest at 30 Hz.

Usually, vibroseis shooting adopts a linear sweep pattern to enhance the energy of weak reflection signals from deep reservoirs by reducing the start sweep frequency. In order to highlight the signal energy at a certain frequency band, a nonlinear vibroseis signal design method is proposed in this paper. Given the time-domain expression for the linear upsweep vibroseis signal is:

$$\mathrm{s}\left(\mathrm{t}\right)=A\sin 2\pi \left(f_1+\frac{\left(f_2-f_1\right)t}{2T}\right)t,$$

(1)

where T is the sweep length; 𝑓₁ and 𝑓₂ are the start and end frequencies; A is the amplitude. The total energy E in the sweep time (Δt) is:

$$E={\sum }_{j=0}^{∆t}{A}^{2}si{n}^{2}2\pi \left(f_1+\frac{\left(f_2-f_1\right)t}{2T}\right)t$$

(2)

Assuming in the above formula that A = 1, 𝑓₁ = 1.5 Hz, 𝑓₂ = 84 Hz, Δt = 2 ms and T = 12 s, the curve of E with t is drawn, see Figure 5.

The straight line in red in Figure 5 is the linear fitting of the energy. It can be seen from Figure 5 that the time integral of the energy is linear approximately, indicating that the distribution of vibroseis energy can be changed and the seismic response of different frequency bands of the target can be improved by changing the sweep time at different frequency bands to achieve the purpose of improving the quality of raw data. Applying this technology to the acquisition of weak signals can further strengthen the energy of low-frequency signals and enhance the anti-shielding ability of signals based on the linear sweep mode of the vibrator. Figure 6 is a comparison of the nonlinear vibroseis sweep signal test records. The start and end sweep signals of the vibrator are 1.5–84 Hz, which is divided into four sections: 1.5–10 Hz, 10–30 Hz, 30–60 Hz and 60–84 Hz by frequency. The sweep time of Signal 1 in the three frequency bands below 60 Hz is 800 ms, 3500 ms and 700 ms longer than that of Signal 2, respectively, while in the 60–84 Hz frequency band, the sweep time of Signal 2 is 5000 ms longer than that of Signal 1. It can be seen from the single-shot records of two different sweep signals and their spectrum curves that the seismic response characteristics of the same target layer are obviously different; the recorded dominant frequency of Signal 1 is lower than that of Signal 2; and the S/N ratio of it is higher. Using this technology, the energy distribution of different vibroseis frequency bands can be changed by changing the sweep time of the target frequency band to achieve the purpose of improving the quality of the data from the target layer.

2.2.2. High-Precision Receiving Technologies

Sensitivity is the corresponding sensitivity of the detector to external excitation (vibration) input, usually expressed as the ratio of output to unit input. It is known from its physical meaning that, under the same conditions, the higher the sensitivity, the stronger the amplitude of the output signal, which is conducive to receiving weak signals. Figure 7 is a comparison of different geophones in the true amplitude of the target layer. In this case, under the conditions whereby other technical indicators are equivalent, a geophone with higher sensitivity has a better response capability to weak signals. Natural frequency is the first element to determine the low cutoff frequency of geophones. The lowest frequency it can receive effectively is one fourth of the natural frequency. To respond to the seismic signals of 1.5 Hz or even lower frequencies, the natural frequency of geophones should not be >6 Hz. Therefore, the geophones with high sensitivity and low natural frequency are required to conduct the seismic survey of deep thermal reservoirs to improve the receiving ability of weak signals. The geophones commonly used currently are shown in

Table 1. Main technical parameters of commonly used geophones (including nodes).

Geophone Type	GT DS5-5 Hz	GT DS5-10 Hz	GT DS- 5 Hz3×1	SN5- 5 Hz	SN5- 10 Hz	SG5- 5 Hz	Smart Solo	Quan- tum	eSeis	30DX -10	30DX- 10 Hz5×2
Natural frequency (Hz)	5	10	5	5	10	5	5	5	5	10	10
DC resistance (Ω)	1920	1800	5760	1820	1550	1850	1850	1850	1920	395	707.5
Damping coefficient	0.6	0.56	0.6	0.7	0.68	0.6	0.6	0.6	0.6	0.3	0.707
Open circuit sensitivity (V/m/s)	83.2	85.8	249.6	86	98	80	80	80	83.2	28	100.5
Distortion	<0.1%	≤0.1%	≤0.1%	≤0.1%	≤0.1%	≤0.1%	<0.1%	≤0.2%	<0.1%	<0.1%	≤0.1%

. For weak seismic signal acquisition, geophones with a natural frequency of 5 Hz and a sensitivity > 80 V/m/s should be selected.

2.2.3. High-Density Acquisition Technologies

In recent years, it has become a consensus in the industry to use high-density 3D seismic survey technologies to improve seismic survey accuracy. High-density acquisition reduces the loss of weak signals and improves the profile and inversion accuracy through sufficient, uniform and symmetrical sampling of the wave field. Bin size and fold are the two most important parameters in high-density acquisition. Small bin size is conducive to improving the spatial sampling accuracy. A high number of folds are conducive to improving the S/N ratio. Both of them are conducive to improving imaging quality.

The advantages of small bin size seismic data can be summarized as follows: (a) the improved spatial sampling accuracy and lateral resolution; (b) a complete linear noise wave field can be obtained, which is conducive to noise suppression; (c) being conducive to the near-surface velocity model of the high-precision of inversion; (d) high-quality first-arrivals can be obtained, which is beneficial to the first-arrival tomographic inversion based static correction. Based on the comparison of the time slices of different bin sizes in Figure 8, the small bin size data shows a clearer fracture system and a higher resolution, which is beneficial to the utilization of the weak signal inside the thermal reservoir.

The high-fold technology is the main technology to improve the S/N ratio and the energy from the target layer in seismic survey. The S/N ratio of the data from the layers inside granite is extremely low due to the influence of many factors such as the low reflection coefficient, poor layering and serious shielding of the layers. High-fold technology is an effective means to solve this problem. Figure 9 shows the illumination effects of the data acquired from granite using different folds. It can be seen from the figure that with the increase in folds, the illumination effect of the target layer is getting better and better. The comparison of the stacking sections of the actual data acquired using different folds (see the red arrows in Figure 10) shows that the higher the folds, the higher the S/N ratio of thermal reservoir data and the clearer the geological phenomenon.

3. The Seismic Acquisition Technology for the HDR in Gonghe Basin and Its Effect

The Gonghe Basin in Qinghai is in the eastern part of the Qinghai–Tibet Plateau, where the resource is sedimentary basin type dry-hot rocks [21,22,23,24]. The thick overlying sedimentary strata inside the basin are Neogene (N) and Quaternary (Q) mudstones and sandstones, and the geothermal resource reservoirs are mainly Indosinian granites, where the granites have good thermal conductivity, the deep heat energy is conducted upward along the granite body, the thermal conductivity of sedimentary formations is extremely low and the overlying sedimentary stratum plays a role of heat insulation, preventing the heat energy from escaping upward continuously, which creates a condition for the formation of high-quality HDR resources [22,23,24,25,26].

The terrain is relatively flat in the study area, where the surface consists of grasslands and sandy hill landforms. The Quaternary sediments in the area are more than 1000 m in thickness, especially the huge poorly sorted gravel layers accumulated at the mountain front on the north and south sides, which are extremely unfavorable for field operation and also have a great impact on the quality of seismic records. In the area, there are mainly sandy gravel pebbles of fluvial facies in the Middle–Late Pleistocene and the granite of the Gonghe Formation in the Early–Middle Pleistocene, the Linxia Formation in the Pliocene, the Xianshuihe Formation in the Miocene and the Middle–Late Triassic. The target for the survey is granite at a burial depth of 1500–6000 m. The seismic survey in the Gonghe Basin was mainly concentrated in the 1990s using vibroseis shooting only with 30–60 folds. Seeing from the previous 2D acquisition sections, the imaging of the HDR mass is unclear (see the position delineated in Figure 11), which cannot meet the precise characterization of the development of faults, fissures and fractured zones. In addition, the energy from deep layers is weak so that the distribution and burial depth of HDR mass is unable to be identified.

According to the data of Well GH1 in the study area (Figure 1), there are great differences in density and velocity between the basement granite and the overlying shale, and a strong wave impedance interface formed between the two. When the seismic incident waves generated by shooting reach this interface, the reflected wave energy generated is strong and the transmitted wave energy is weak so that the shielding effect is strong, resulting in the weak reflection energy from the target layer, which affects the imaging of the target layer. The difference in bedrock granite formation is small in velocity and density and the reflection coefficient is low, which is unfavorable for seismic imaging.

In order to acquire weak seismic signals from thermal reservoirs during the 3D seismic acquisition of HDR in the Gonghe Basin, and to improve the data imaging effect, the following measures were implemented In terms of shooting, first, a large-tonnage EV56 high-precision low-frequency vibrator array (2 vibrators × 1 sweep) is used for shooting, while increasing the sweep length (18 s) to enhance the ability of seismic waves to penetrate the top interface of igneous rocks and improve the reflected energy of weak signals. Second, the broadband sweep technology is used to reduce the start frequency of sweep and the 2–96 Hz start and end frequency of sweep are used—low frequency helps to overcome the shielding effect of the top granite interface and improve the down transmission of energy. High frequency helps to improve the resolution. In terms of receiving, a single smartsolo-5 Hz low-frequency node is used for receiving, which, on one hand, avoids the array effect from damaging weak signals. On the other hand, the low natural frequency (5 Hz) and high sensitivity (80 V/m/s) help to improve the receiving of weak low-frequency signals. In terms of geometry, the high-density acquisition technology (the geometry parameters: small bin size 10 m × 10 m, ≥600 folds (

Table 2. The 3D Seismic Geometry in Well GH1 Area, Gonghe Basin.

Geometry Type	14L × 15S × 253T Orthogonal Full Spread Receiving	Source Line Interval	60 m/120 m
Bin size	10 m × 10 m	Source point interval	10 m/20 m
Folds	600–1140 folds	Max crossline interval	3890 m
Receiving traces	3308 traces	Max offset	6374 m
Receiver line interval	300 m	Aspect ratio	0.77
Group interval	20 m

), the coverage density up to 6 million/km²) is applied, which helps to improve the imaging of weak signals from thermal reservoirs.

Figure 12 is the stacking section of the HDR in the Gonghe Basin, in which the Cenozoic sedimentary stratum shows obvious wave resistance characteristics, the top surface of the granite has clear characteristics and there is rich information inside the deep granite, the survey demands of HDR are achieved.

4. Conclusions

(1) There are three main causes for the weakness of the seismic reflection signals from deep thermal reservoirs: (a) There are large differences in velocity and density, and large wave impedance differences between the thermal reservoir and the caprock so that a strong wave impedance interface, which has a strong shielding effect on the down transmission of energy, forms; (b) The velocity and density have little changes inside the thermal reservoir so that no effective wave impedance interface is available, resulting in weak reflected energy; (c) There are high-temperature fluids such as hydrothermal fluid and steam in the thermal reservoir so that the absorption and attenuation are serious.

(2) The weak signal acquired in the seismic survey of deep thermal reservoirs is a difficulty, the processing and interpretation of weak signals are also challenging. Seismic survey is a systematic work. For the seismic survey projects of deep thermal reservoirs, it is very important to handle the integrative design of acquisition, processing and interpretation properly when conducting seismic data acquisition.

(3) When solving the deep thermal reservoir weak signal acquisition problem, three aspects should be considered comprehensively: shooting, receiving and geometry through the application of a series of supporting technologies. For shooting, through the use of the high-energy low-frequency shooting technology, the penetration ability and down transmission of energy of seismic signals can be improved; for receiving, the weak signal receiving capacity can be improved by using single-point, high-sensitivity geophones; in terms of geometry, the energy and the S/N ratio of weak signals can be improved by increasing the folds thanks to the use of the small binsize, high-fold acquisition technologies. The technologies are applied in the seismic acquisition of dry-hot rocks (HDR) in Gonghe Basin, Qinghai. The Cenozoic sedimentary stratum shows obvious wave resistance characteristics, the top surface of the granite has clear characteristics and rich inner information of the deep granite is available. The survey demands of HDR are achieved.