Geochemical Characteristics and Origin of Nuanquanzi Geothermal Water in Yudaokou, Chengde, Hebei, North China

0. INTRODUCTION

In the context of carbon neutralization, the utilization of geothermal resources has increased significantly over the past few years owing to its advantages of being clean, renewable, and environmentally friendly with low carbon emissions (Zhang and Hu, 2018; Zhu et al., 2015). The Chengde City in northern Hebei Province has been termed the "Thermal River", and obtained broad application prospects of clean energy due to its abundant geothermal resources (Yang et al., 2012). Geothermal systems in North China can be divided into deep geothermal systems along the North China fault zone and thermal spring systems in the North China intracontinental orogenic belt (Chen et al., 1990). However, previous geothermal studies were predominantly concentrated in Beijing, Tianjin, and Xiong'an in the North China fault zone (Wang et al., 2020, 2017; Zhang et al., 2019); few studies were focused on the geothermal system in the inland orogenic belt of North China.

The Yudaokou area is located on the northwestern margin of the North China rift basin and the Yanshan uplift belt on the southern margin of the Bashang Plateau in northern Hebei Province. The spatial distribution of large-scale deep faults developed during the Early Yanshanian movement generally extends up to hundreds of kilometers; some of these faults penetrate the silicon-magnesium layer or upper mantle (Liu et al., 2019; Jiang et al., 2013; Davis et al., 2001), which plays an essential role in the upward conduction of deep geothermal groundwater. The crustal deformation, mantle upwelling, earthquake, and magmatic activity caused by the destruction of the North China Craton provide favorable conditions for the heat migration, and geothermal water circulation from the deep part of the North China fault basin to the northwest (Fig. 1) (Zhu et al., 2012; Ren et al., 2002). Study on the Nuanquanzi geothermal field of Yudaokou Ranch in the Yanshan uplift is of great significance for understanding the formation mechanism, occurrence environment and circulation mechanism of geothermal fluid in the inland orogenic belt of North China, at the same time, it is helpful to clarify those difference and correlation with the geothermal system in the North China fault depression basin.

Figure  1.  Geological map, regional geological structure background, and geothermal and non-thermal water distribution in Yudaokou, and thermal springs in Chengde City (map of China after GS(2019)1696). 1. Alluvial marsh sand gravel, sandy loam; 2. alluvial sandy loam; 3. eolian sand; 4. Hannuoba Formation basalt; 5. Yixian Formation andesite, tuff; 6. Zhangjiakou Formation rhyolite; 7. geothermal well; 8. river water sample point; 9. well; 10. spring; 11. deep fault; 12. hot spring; 13. radon measurement area; 14. county and city names; 15. electromagnetic earth electromagnetic profile of controllable source; 16. fault; 17. speculative fault; 18. river; Ⅰ₁. Inner Mongolia-Greater Khingan Range fold system; Ⅱ₁¹. Inner Mongolia Waresi late fold belt; Ⅱ₂¹. China-Korean quasi-platform Inner Mongolian Axis; Ⅲ₁¹. Doron anticlinorium; Ⅲ₂¹. Weichang arch break beam; Ⅲ₂². Chengde arch break beam; Ⅲ₂³. Malanyu. F1. Kangbao-Weichang fault; F2. Fengning-Longhua fault; F3. Damiao-Niangniangmiao fault; F4. Shangyi-Pingquan fault; F5. Huangqi-Wulonggou fault; F6: Pingfang-Sangyuan fault; F7. Xinglong-Xifengkou fault.

DownLoad: Full-Size Img PowerPoint

In this study, the formation mechanisms, origin, circulation depth, and geothermal reservoir temperature of the Nuanquanzi geothermal system were elucidated by classical hydrogeochemical analysis, multi-isotopes approach (δD, δ¹⁸O, δ¹³C, δ⁸⁷Sr/⁸⁶Sr), groundwater ¹⁴C dating by accelerator mass spectrometer (AMS), and integrated geophysical prospecting. Surface-soil radon gas measurements and controllable source audio-frequency magnetotelluric (CSAMT) geophysical inversion was applied to detect the development of fault structures associated with the origin of geothermal water. In summary, this research on the geochemical characteristics and origin of Nuanquanzi geothermal water plays pivotal roles in providing a scientific basis and guidance for the development and utilization of geothermal resources in the inland orogenic belt of North China.

1. GEOLOGICAL SETTING

1.1 Regional Geology

The Nuanquanzi area is located in northern Weichang County, Chengde City. The regional tectonic division is located in the folding unit, spanning from the platform of China and Korea to the margin of the North China Platform and connected with the fold system of the Greater Khingan Range in Inner Mongolia, namely the tectonic unit of the Qipanshan depression in the Duolun anticlinorium. In the tectonic movement history, the study area experienced various stages of movement in Yanshan, Himalayas, and neotectonics. In the early stage, owing to the influence of squeezing and welding between the Siberian Plate, Tarim-North China Plate, and South China Plate, large-scale faults and folds spreading from east to west were formed and represented by the Kangbao-Waichang deep fault (F7, Fig. 1). Subsequently, owing to the tectonic movement in the Pacific Rim, the fault and folding system trending in the northeast to north-northeast direction developed and superimposed with the existing E-W faults, as represented by the upper Huangqi-Wulonggou deep fault (F7, Fig. 1) (Liu et al., 2019; Davis et al., 2001). The Nuanquanzi area is located at the intersection of two deep and secondary faults; it has developed an inherited northeast-trending hidden fault. The strata exposed in the study area mainly comprise Neogene, Cretaceous, and Jurassic volcanic rocks (Figs. 1 and 2). The underlying bedrock includes the Jurassic tuffaceous sands and mudstone, Mesozoic–Cretaceous volcanic-sedimentary rock combination of the Dabeigou Formation (K₁d), and sandstone of the Qingshibian Formation (K₁q).

Figure  2.  Formation lithology and the measured gradient curve of ground temperature from DW01. Q. Quaternary sandy loam, fine sand with sand gravel; N₁h. basalt of Hannuoba Formation; K₁y¹. andesite intercalated with amygdaloid and basaltic andesite; K₁y². Andesitic breccia tuff with volcanic breccia; J₃zh¹. tuffaceous sandstone intercalated with sandy conglomerate and mudstone; J₃zh². sandstone conglomerate and mudstone.

DownLoad: Full-Size Img PowerPoint

1.2 Hydrogeology and Geothermal Setting

Previous studies have demonstrated that the terrestrial heat flow value in the Yanshan uplift zone is relatively low, with an average value of 42.5 mW/m² and a varied range of 25.5–61.0 mW/m² (Yang et al., 2012; Wang, 2002). The terrestrial heat flow value of the whole Chengde is 30.1 mW/m², which is relatively lower than the average value of 51.5 mW/m² of North China, and the measured value of Qijia-Maojingba geothermal field is 74.9 mW/m². The terrestrial heat flow value in the mountainous area is lower than that in the plain area of North China, for instance, the Xiong'an New Area (Zhu et al., 2021; Liu et al., 2020). According to the measured data of the Beijing-Tianjin-Hebei area, the underlying bedrock of Yudao-kou is mainly sandstone and mudstone with an average thermal conductivity of 1.72 W/(m·K). The thermal conductivity of sandstone-siltstone in the adjacent regions in Wuqing and Jixian in Tianjin City is 1.54–1.72 W/(m·K) (Ruan et al., 2017; Wang et al., 2017). The thermal conductivity of rocks decreases gradually with the formation of the new North China fault basin geothermal system. The caprocks of the Yudaokou geothermal field include mudstone, argillaceous sandstone and its conglomerates, tuffaceous sands, and tuff of the Cretaceous and Neogene. Significantly, the argillaceous sedimentary rock of the Zhangjiakou Formation with poor water permeability and low thermal conductivity was widely distributed, which plays an essential role in water and thermal insulation.

There are six natural cold springs and low-medium thermal springs, and two geothermal drilling wells in the study area. The DW01 geothermal well is located in Nuanquanzi Village, and DW02 geothermal well situated beside Yuke Road on the south side of Bashang Village (Fig. 1). The depth of DW01 is 1 973.6 m, and the flow rate of the artesian well is approximately 160 m³/d. The artesian outlet water temperature of the DW01 wellhead is 48 ℃, and that at the bottom is 61.06 ℃ (Fig. 2). The depth of DW02 is 1 300 m, and the flow rate is approximately 500 m³/d with an outlet water temperature of 37 ℃ and a bottom temperature of 41 ℃. The depth of the constant temperature zone in the study area is approximately 15 m, and the continuous temperature zone has an average annual temperature of 4.0 ℃. According to the temperature measurement curve (Fig. 2), the average geothermal gradient of DW01 is 2.87 ℃/100 m, and that of DW02 is 2.81 ℃/100 m. The average temperature gradient in this area is consistent with the interval value of the geothermal gradient in the Jizhong sag from 2.0 to 3.0 ℃/100 m (Wang et al., 1993), while being slightly lower than the value of 3.2 ℃/100 m in the central part of the North China fault basin (Wang et al., 2017). The temperature measuring curve (Fig. 2) shows a linear trend, indicating that the heat transfer process of the formation in the study area is mainly conduction. The geothermal gradient below 1 300 m is relatively more prominent due to the variation in the ground temperature field under the action of rising water flow; meanwhile, the heat conduction efficiency increases with a rise in the compaction and stress of the sediment layer, and the rock radiation and thermal conductivity of some special lithology (feldspar) increase under high-temperature conditions (Lee and Deming, 1998).

2. SAMPLING AND ANALYSIS METHODS

2.1 Hydrochemical Analysis

A total of 27 water samples were collected from the Yu-daokou area, and 21 geothermal well water and thermal spring samples from the surrounding fault zone of Chengde were collected from previous studies. Among water samples in Yudao-kou, there were geothermal well sample (n = 3), cold spring water (n = 6), shallow groundwater (n = 6), Xiaoluan River water (n = 10), and precipitation sample (n = 2). And typical geothermal water and thermal spring samples taken from the surrounding area was shown in Fig. 1 and Table 1.

Table  1.  Chemical and isotope compositions of water samples from Yudaokou and thermal springs collected in adjacent areas in Chengde, North China

No.	Sample type	Sampling location	pH	T		TDS	Na+	K+	Ca2+	Mg2+	Cl-	SO42-	HCO3-	CO2	SiO2	F-	H3BO3		δD	δ18O	δ87Sr/86Sr
				(℃)		(mg/L)													(‰)
DR01	Geothermal water	Nuanquanzi (DW01 wellhead artesian water)	8.26	42.0		1 215.89	453.80	2.38	1.64	0.49	44.55	26.03	1 078.00	148.00	25.86	29.69	1.08		-97.11	-13.75	0.707 8
DR02		Nuanquanzi (DW01 wellbottom water)	8.54	48.0		1 350.00	796.00	2.85	3.60	0.10	23.7	48.50	1 635.00	72.00	32.60	31.80	4.67		—	—	—
DR03		Bashang geothermal well (DW02)	8.23	37.0		1 670.62	682.50	2.16	2.40	0.12	36.64	15.93	1 657.76	102.00	20.00	31.60	2.15		-98.86	-14.03	0.707 6
YS01	Typical cold spring in Yudaokou	Nuanquanzi	8.78	6.9		152.21	2.80	3.22	19.73	5.84	3.79	15.86	64.93	0	20.79	0.13	< 0.01		—	—	—
YS02		Hounuanquanzi	8.28	7.0		196.34	59.03	3.15	15.98	2.83	3.77	11.75	70.18	65.85	30.50	0.13	0.04		-80.92	-11.25	0.710 1
YS03		North Hounuanquanzi	7.44	7.8		244.59	2.16	1.05	37.41	7.74	3.70	6.46	152.16	0	24.10	0.28	< 0.01		—	—	—
YS04		North Hounuanquanzi	6.90	6.7		155.09	7.51	0.65	21.16	3.58	2.95	6.88	89.83	2.07	25.70	0.23	< 0.01		-82.31	-11.13	—
JS01	Precipitation	Precipitation	6.40	7.0		11.38	0.26	0.37	2.01	0.19	0.41	2.35	5.64	2.07	0.16	0.01	ND		-41.07	-8.17	—
SW01	Thermal springs of Chengde	Shanwanzi geothermal water	8.10	82.0		763.94	92.90	6.20	85.01	2.41	38.29	31.08	447.46	5.43	42.63	10.31	< 0.01		-94.80	-11.70
SW02		Shanwanzi thermal spring 1	8.30	74.5		797.00	214.00	6.06	5.00	0.60	34.70	33.60	423.00	0.00	11.96	1.45	< 0.01		-91.71	-12.80
SW03		Shanwanzi thermal spring 2	7.10	72.0		368.15	15.34	2.64	58.34	18.90	18.44	8.24	216.87	0	20.8	0.38	< 0.01		—	—
SW04		Shanwanzi thermal spring 2	7.30	68.0		399.92	17.81	2.76	45.34	25.43	24.11	9.47	231.83	0	19.76	1.25	< 0.01		—	—
QJ01		Qijia thermal spring 1	7.86	42.0		604.00	157.20	4.32	12.65	0.55	27.28	199.20	134.20	2.20	120.20	13.40	< 0.01		-84.80	-11.00
QJ02		Qijia thermal spring 2	8.32	97.0		651.00	166.70	5.24	12.47	0.18	29.02	218.50	106.20	0	123.50	14.40	< 0.01		—	—
MS01		Maojingba thermal spring	8.57	71.5		668.00	18.40	10.00	11.60	0.50	31.90	234.00	104.00	18.00	74.80	15.80	0.30		-83.30	-9.80
HT01		Hongtangsi thermal spring	8.88	49.5		302.00	98.90	2.17	5.00	0.60	19.90	75.90	63.50	0	75.70	17.50	0.20		-95.00	-12.60
BS01		Beidaba thermal spring	8.18	35.5		757.00	208.00	3.69	5.00	0.60	42.50	250.00	171.00	6.00	64.40	13.00	0.26		-86.50	-11.80
BH01		Beiwenquan thermal spring	8.08	68.0		832.00	234.00	6.20	5.00	0.60	35.40	134.00	354.00	6.00	44.30	22.50	0.64		-90.90	-11.20
NH01		Nanwenquan thermal spring	7.68	49.8		826.00	224.00	5.37	11.40	0.60	37.20	242.00	248.00	0.00	56.30	14.00	0.60		-88.80	-10.90

 | Show Table

DownLoad: CSV

The water samples were collected and conserved using HPET plastic bottles and thermo glass bottles of different specifications, sealed by parafilm, and stored in a portable low-temperature refrigerator (4 ℃). Moreover, the water samples for determining isotopes (δD, δ¹⁸O, δ¹³C, δ⁸⁷Sr/⁸⁶Sr) were filtered through 0.45 μm membranes and subsequently separated in different aliquots. The in-situ physicochemical parameters of samples, including temperature (T), pH, conductivity (EC), dissolved oxygen (DO), and redox potential (ORP), were determined by the portable multi-parameter water quality rapid detector (Multi-3430 Germany) during field sampling.

The main chemical ions of K⁺, Na⁺, Ca²⁺ and Mg²⁺ were analyzed using inductively coupled plasma optical emission spectrometry (Agilent ICP-97). The main anions of SO₄^2- and Cl^- were determined using ion chromatography (IC) (Dionex ICS-1100), and F^- were determined using ion-selective electrode with an accuracy of 0.05 mg·L^-1. The analyses conducted by both instruments were within a relative standard deviation of 5%. Free CO₂ and HCO₃^- were measured using manual hydrochloric acid titration. The soluble SiO₂ was measured using a spectrophotometer (UV2900), and HBO₃ was determined using curcumin spectrophotometry, both with an accuracy of 0.10 mg·L^-1. The δD, δ¹⁸O, δ¹³C_DIC and ¹⁴C groundwater dating were measured at the Laboratory of Beta Analytic Inc. using an accelerator mass spectrometer (AMS) coupled to an isotope ratio mass spectrometer (IRMS, Thermo Fisher Delta V Advantage) with a precision of ±0.1‰, ±0.2‰, ±0.3‰, respectively. The δ⁸⁷Sr/⁸⁶Sr were determined using multi receiver inductively coupled plasma mass spectrometer (MC-ICP-MS) (NU Plasma Ⅱ) with an accuracy of 0.004‰ (SD 1s) in China University of Geosciences (Wuhan).

The addition ratio of blank samples, duplicate samples, and standard reference materials was 10% to ensure quality control during the laboratory analysis. The calibrated water apparent age was reported as RCYBP (radiocarbon years before present; "present" = AD 1950). According to international convention, the modern reference standard is 95% of the ¹⁴C activity of the National Institute of Standards and Technology Oxalic Acid (SRM 4990C) and calculated using the Libby ¹⁴C half-life (5 568 years). The DIC extraction consisted of injecting sample water into an acid bath attached to an evacuated collection line. pH was reduced to < 1 and evolved CO₂ was dried with methanol slush and collected in liquid nitrogen. CO₂ was then graphitized over cobalt in a hydrogen atmosphere to produce the target for the AMS. Agreement between expected and measured values is taken as being within two standard variance (2σ) agreement to account for total laboratory error, and the laboratory conducts pretreatment, graphitization and AMS analysis on a same single sample for 2–3 times, to reduce the standard error of dating within ±30 a BP.

2.2 Geophysical Exploration Methods

Geophysical exploration of the fault structure was performed using the controllable source audio magnetotelluric method (CSAMT) owing to its strong anti-interference ability. The V8 multifunctional electrical method acquisition system of the Canadian Phoenix Company was used. The detection frequency receiving range of the instrument was 10 000–0.000 05 Hz. Two CSAMT geophysical sections with a total of 124 points (distance between each two points is 50 m) were taken in the study area. The acquisition frequency of the instrument in these two profiles was 1–8 192 Hz, and the effective detection depth was more than 1 000 m. The received data show relatively considerable quality with the relative mean square error of resistivity being ±6.03%, while the relative mean square error associated with the impedance phase at ±9.35%.

The soil radon gas was determined using a radon measurement instrument (HDC-C type of Shijiazhuang Nuclear Industrial China). Soil radon was measured for a total area of 21 km², with the distance between exploration lines, and two adjacent points being 100 and 40 m, respectively. A total of 57 measuring lines and 4 577 measuring points were completed to delineate the radon gas concentration site. The sensitivity of the soil radon concentration measurement was greater than 300 Bq/m³. The depth of gas extraction at each point was generally 0.7 to 1.0 m, and the number of air extractions at each end was generally 6–10. In addition, the re-detection rate of the measuring point for quality control was 5.8%.

At the same time, soil radon was also measured on two CSAMT profiles to comprehensively indicate the potential fault structure through the CSAMT bathymetric inversion of resistivity and distribution change characteristics of radon gas; this was achieved by plotting the background value R_nB and abnormal threshold R_nF on the curve of radon concentration. In addition, a 177.8 mm diameter oil casing was drilled and installed for geophysical logging to test the water temperature, and well deviation after the DW01 whole well drilling was completed.

3. RESULTS AND DISCUSSION

3.1 Hydrogeochemistry of Groundwater

The temperature of Nuanquanzi geothermal water varied from 37.0 to 48.0 ℃, and the pH value changed from 8.23 to 8.54. The total dissolved solids (TDS) concentration of Nuanquanzi thermal water varied from 1 350.00 to 1 670.62 mg/L. The principal cation and anion in the thermal water were Na⁺ and HCO₃^-, respectively. Thus, the type of hydrochemistry is primarily HCO₃-Na. The temperature of non-thermal water in the shallow groundwater varied from 4.9 to 28.5 ℃, and the pH ranged from 6.88 to 7.86. The average TDS concentration was 189.49 mg/L, thereby representing freshwater with low salinity. The hydrochemical types of shallow non-thermal groundwater were mainly HCO₃-Na·Ca and HCO₃-Ca·Mg. The temperature of the geothermal water in adjacent areas of Chengde varied from 35.5 to 97.0 ℃ with the pH value varied from 7.10 to 8.88; and the TDS concentration ranged from 302 to 832 mg/L. The hydrochemical types of these geothermal waters were mainly HCO₃·SO₄-Na and SO₄-Na (Fig. 3).

Figure  3.  Piper (1944) diagram for all water samples from Yudaokou and its adjacent areas in Chengde, North China.

DownLoad: Full-Size Img PowerPoint

The hydrochemical type of the Shanwanzi and Nuanquanzi geothermal water was both HCO₃-Na (Fig. 3), which were characterized as open geothermal water. The concentration of Cl^- in the Shanwanzi and Nuanquanzi geothermal water was much lower than that of adjacent thermal fields in the North China depression basin. The hydrochemical composition of groundwater was derived from water-rock interaction during the runoff process; the longer the runoff path of geothermal water, the higher the TDS (Daniele et al., 2020). The TDS concentration of geothermal water samples in Nuanquanzi and Shanwanzi was significantly higher, and the SO₄^2- concentration (5.93 to 48.50 mg/L) was relatively lower than that of samples from the Longhua-Fengning area (165.3 to 387.2 mg/L), indicating that the geothermal water in the Nuanquanzi-Shanwanzi area had a longer runoff path and enormous circulation depth. The formation environment of thermal springs in the Longhua and Fengning areas is a relatively closed reduction environment (Zhang, 2012).

According to the relationship between Cl-SO₄-HCO₃ concentration, geothermal water can be classified into four categories: deep mature water that is rich in Cl^- and originates from deep environments; SO₄^2- rich steam water derived from deep geothermal steam fluid; HCO₃^- rich peripheral water, which is dominated by the infiltration of atmospheric precipitation; and Cl^--SO₄^2- volcanic water produced by mixing deep geothermal water with steam condensation (Giggenbach et al., 1988). According to Fig. 4, the geothermal water in Nuanquanzi is categorized as peripheral water. The recharging source of geothermal water is dominated by atmospheric precipitation and is weakly influenced by deep geothermal water (Baba et al., 2019). The other geothermal water in adjacent areas falls on the mixed Cl^--SO₄^2- water volcanic condensates and the steam-heated water terminal in Fig. 4, thereby suggesting that the geothermal spring water originates from deep geothermal steam fluid heating and condensation mixing.

Figure  4.  Cl-SO₄-HCO₃ triangular diagram of water samples from Yudaokou and its adjacent areas (based on Nicholson, 1993).

DownLoad: Full-Size Img PowerPoint

Trace elements in geothermal water mainly originate from water-rock interactions or deep geothermal fluid mixing during the runoff process (Stefánsson et al., 2019; Guo and Wang, 2012). The contribution of magmatic degassing process of high-temperature geothermal system to Cl^- concentration in geothermal water is much higher than that of water-rock interactions (Noble et al., 1967), and Cl^- is difficult to be adsorbed by surrounding rock minerals or precipitated in the form of hydrothermal altered minerals during the upwelling of geothermal fluid (Truesdell et al., 1977). Therefore, as a conservative component, Cl^- was used to indicate the mixing process of geothermal fluids. In additional, the enrichment of F^- in geothermal water was dominated by the water-rock interaction of the dissolution of fluorine-containing minerals. Thus, the circulation path of thermal water can be identified through the characteristic relationship of Cl^-, and F^- between other hydrochemical components in thermal water (Alçiçek et al., 2016; Li et al., 2015).

The F^- concentration of Nuanquanzi thermal water varied from 29.69 to 31.80 mg/L, which is significantly higher than that in shallow groundwater (average 0.28 mg/L) and other geothermal water (average 9.68 mg/L) in adjacent areas. The F^- concentration in geothermal water generally tends to increase with the temperature as shown in Fig. 7a. The solubility of fluorine-containing minerals in surrounding rock increases with the groundwater temperature, resulting in the enrichment of F^- in groundwater (Yuan et al., 2021). Moreover, the pH value of high fluorine geothermal water was generally greater than 7.5 (Fig. 7b), and the F^- concentration generally increases with the pH value. The alkaline water environment is conducive to the dissolution of fluorine containing minerals or promote the desorption of F^- adsorbed on the mineral surface into water (Jacks et al., 2005). In addition, the larger the thermal water circulation depth, the higher the F^- concentration in the geothermal water in the same geothermal field, in this way, the geothermal water with high F^- concentration demonstrated to have a longer circulation path (Yuan et al., 2021). The igneous rocks in the study area are rich in biotite, amphibole, fluorite, and associated mineral of fluoroapatite. The long-term water-rock interaction led to the release of F^- from fluorine-containing minerals into the water. The F^- concentration of Nuanquanzi thermal water was relatively higher than those thermal water in adjacent areas, indicating its circulation path longer than the southern thermal spring system. On the other hand, the mixing process with the rising fluorine-rich deep crustal thermal water through faults and fracture zones also affects the F^- concentration in shallow geothermal water (Guo and Wang, 2012). The temperature of Nuanquanzi thermal water was lower, but F^- concentration significantly higher than those samples of the southern thermal springs, suggesting that the mixing effect of cold water in Nuanquanzi thermal reservoir was relatively weaker than that of southern thermal spring system.

The concentration of soluble SiO₂ in the Nuanquanzi and Shanwanzi geothermal waters was generally higher than that of shallow groundwater but lower than that of the southern thermal springs (Fig. 5c). A possible reason for this phenomenon is that the terrestrial heat flow and thermal reservoir temperature gradually decreased from the center to the northern margin of the North China rift basin in the Yanshan mountainous area (Liu et al., 2020; Yang et al., 2012). Simultaneously, the Cl^-/SiO₂ ratio of Nuanquanzi and Shanwanzi geothermal water was higher than that of the southern thermal springs, suggesting a longer residence time, sufficient water-rock interaction, and a deeper circulation path, matching well to indicative characteristics of fluorine concentration.

Figure  5.  Relationship between fluoride, chlorine, and trace elements concentration in groundwater from Yudaokou and its adjacent areas.

DownLoad: Full-Size Img PowerPoint

Volcanic clastic rock formations are widely distributed on both sides of the Kangbao-Weichang deep fault. The Sr²⁺ in the groundwater mainly originated from the deep circulation leaching process of the rock minerals. The high-temperature thermal reservoir accelerated the dissolution rate of minerals in the geothermal fluid in the fault zone. Thus, the Sr²⁺ concentration in the thermal springs around the fault zone was higher than that of the Nuanquanzi-Shanwanzi geothermal water. The Sr²⁺ and Cl^- concentrations of the DR01 and DR03 artesian geothermal water samples at different depths were similar, which suggests that they likely originated from the same confined aquifer. Li et al. (2011) distinguished the formation and evolution of the stratum environment of hydrogeochemistry by clarifying the relationship between Mn²⁺ and Ca²⁺ concentrations in geothermal water. The thermal water may have flowed through sandstone, slate, and mudstone distribution areas when it had a high Mn²⁺ concentration. Both the Mn²⁺ and Ca²⁺ concentrations of the Nuanquanzi geothermal waters were lower than those of the non-thermal water samples (Fig. 5e), indicating that the recharged groundwater flows through the magmatic and metamorphic aquifer, which is consistent with the wide distribution of the magmatic rocks around the fault zone and that the pyroclastic rocks are commonly distributed in the spring runoff area. It is generally believed that boron contained in geothermal fluid is contributed by primary magmatic water (Ellis and Mahon, 1964). Overall, the Ca²⁺ concentration of geothermal water was generally lower, and the Sr²⁺, and H₃BO₃ concentrations were higher than those of non-thermal water (Figs. 5d, 5e, 5f). The ratios of Sr²⁺/Ca²⁺ and B/Cl followed the order of geothermal water > thermal spring > shallow groundwater. The Nuanquanzi geothermal waters in northern Chengde City have exhibited more sufficient water-rock interaction, slower circulation rate, longer residence time, and deeper circulation depth than that seen in shallow groundwater and other thermal water sources in central and southern Chengde.

The relationship between the groundwater ion coefficients, such as sodium chloride coefficient (rNa/rCl), desulphite coefficient (rSO₄ × 100/rCl), metamorphic coefficient (r(Cl-Na)/rMg), and TDS can be used to clarify the depth of groundwater circulation and the sealing degree of the hydrogeochemical environment (Mao et al., 2021; Raiber et al., 2009). The sodium chloride coefficient of the DR02 sample at the bottom of the Nuanquanzi geothermal well (51.84) was significantly higher than those of DR01 (15.72) and DR03 (28.75) samples at the artesian thermal water of the wellhead (Fig. 6a). The sodium chloride coefficient of thermal springs in adjacent areas varied from 0.89 to 10.20, with an average value of 6.13; the average sodium chloride coefficient of shallow groundwater was 3.72. The sodium chloride coefficient and TDS value exhibited the following order: geothermal water > thermal spring > shallow groundwater. The desulphite coefficient of the thermal springs varied from 29.05 to 369.41, while the shallow groundwater varied from 0.71 to 309.49, with an average value of 138.18; both these ranges of values were higher than the desulphite coefficient of the Nuanquanzi geothermal water (Fig. 6b). The average value of the metamorphic coefficient of the different samples exhibited the following order: geothermal water > thermal spring > shallow groundwater. The average values of the metamorphic coefficients of shallow groundwater and thermal springs were 2.58 and 117.47, respectively (Fig. 6c). In general, the sodium chloride and metamorphic coefficients of groundwater in the study area increased with the TDS concentration. The sodium chloride and metamorphic coefficients of Nuanquanzi thermal water were higher than those of other thermal springs, indicating that the sealing degree of the hydrogeochemical formation environment was higher and that the residence time was longer in central and southern Chengde than that in the thermal water. The desulphite coefficient of the thermal springs was the highest, owing to the fact that the thermal springs were mainly exposed in the water-conducting fault zone. The runoff paths were mostly magmatic and pyroclastic. The mixing of deep geothermal steam condensation and infiltration of cold water was intense and conducive to the formation of volcanic geothermal water comprising Cl^- and SO₄^2-.

Figure  6.  Relationship between total dissolved solids and sodium chloride (a), desulphite (b), and metamorphic coefficients (c) of different water sample types from Yudaokou and its adjacent areas.

DownLoad: Full-Size Img PowerPoint

3.2 Origin of the Groundwater

3.2.1   Hydrogeochemistry isotopic characteristics

The relationships of δD and δ¹⁸O were illustrated in Fig. 7a together with the local meteoric water line (LMWL), which was determined by the atmospheric precipitation line in Chengde City (Sun et al., 2020). The isotopic compositions of δD and δ¹⁸O at DR01 were -97.11‰ and -13.75‰, respectively; those at DR03 were -98.86‰ and -14.03‰, respectively (Table 1). Both these sets of values were lower than those of the shallow groundwater and thermal springs in the local area. The shallow groundwater samples were almost identical and closely focused on the global meteoric water line (GMWL), indicating that the primary groundwater recharge source was offered by the infiltration of atmospheric precipitation. The geothermal springs in adjacent areas were also close to the GMWL but exhibited deviation to the right of the LMWL (Fig. 7a), which suggests that the geothermal springs were considered to be of meteoric origin and that isotopic shift of oxygen was caused due to relatively high-temperature water-rock interactions. The Nuanquanzi geothermal water was shifted to the left of the LMWL (Fig. 7a), indicating that its formation was achieved in a medium-low temperature thermal reservoir environment.

Figure  7.  Relationship among δD and δ¹⁸O (a) and δ D and Cl^- (b) for thermal and non-thermal water from Yudaokou and its adjacent areas. The straight line (δD = 8.0 δ¹⁸O + 10.0) (Craig, 1961) represents the global meteoric water line (GMWL), and the dashed line (δD = 9.37 δ¹⁸O + 28.73) represents the local meteoric water line (LMWL) (based on Sun et al., 2020). The isotopic geoindicator for hydrological and water-rock interaction processes: 1. subduction volcanic steam; 2. magmatic arc; 3. felsic crust magma; 4. magmatic melt degassing; and 5. mid-ocean ridge basalt in plot (a).

DownLoad: Full-Size Img PowerPoint

The δD value generally decreased with the depth of groundwater recharge. The δD values of geothermal water in the local area were lower than the average value of -85.95‰ for thermal springs in adjacent regions of Chengde, and average value of -79.22‰ of geothermal water in the central and northern parts of the North China fault basin (Zhang et al., 2019; Pang et al., 2018). That is, the source and circulation depth of geothermal water in the Yanshan uplift belt were deeper than those in the North China fault zones. The isotopic compositions of δD and δ¹⁸O of Nuanquanzi geothermal water were relatively depleted in the thermal springs and shallow surface water, which was related to the CO₂ exchange with basalt and magmatic mantle minerals through water-rock interaction and elevation effects during groundwater recharge (Zhang et al., 2019; Pang et al., 2018). The Inner Mongolia Plateau and the Bashang Plateau in the north of the study area were considered to be the recharge area of the geothermal system of Nuanquanzi and Shanwanzi because the topography of these areas was higher than that of the surrounding areas.

The δD value generally decreases as the Cl^- concentration increases in geothermal water. Thus, the mixing of geothermal water in shallow aquifers can be identified by evaluating the relationship between δD and Cl^- concentrations in groundwater (Pürschel et al., 2013). To account for the δD-Cl^- value of high-temperature geothermal fluids (Li et al., 2018), the Nuanquanzi thermal reservoir was considered to have been mixed in deep aquifers with the CO₂ condensate (Fig. 7b). Therefore, the Nuanquanzi geothermal water was rich in CO₂ due to degassing and overflow of the magma melting, with the free CO₂ concentration of water samples varying from 72 to 148 mg/L (Table 1). The thermal springs in adjacent areas were impacted more by mixing with the infiltrating groundwater of shallow aquifers.

In general, dissolved inorganic carbon (DIC) in groundwater obtain the following sources: soil CO₂, carbonate mineral dissolution, direct input of precipitation, CO₂ exchange with the atmosphere, and deep mantle CO₂. Numerous studies suggest that the δ¹³C of soil organic matter dominated by C3 plants ranges from -30‰ to -24‰ (Balesdent et al., 1993). The δ¹³C of inorganic carbon dioxide generally distributed between -8‰ and +3‰. The δ¹³C value of modern atmospheric CO₂ ranges from -7‰ to -8.5‰ (Alçiçek et al., 2018). The δ¹³C value of carbon dioxide derived from carbonate metamorphism is close to that of carbonate rocks, which is approximately 0 ± 3‰ and -6‰ ± 2‰ of the carbon dioxide derived from volcanic magma and mantle among the inorganic carbon dioxide sources (Sano and Marty, 1995). Concurrently, during the metasomatism of subducted carbonate and silicate melts on the crust mantle boundary and the exchange between basalt and igneous carbonate in the North China Craton (Chen et al., 2017), calcite dissolution precipitation was involved in the carbon isotope fractionation process of groundwater because of the carbonated sediment recycling.

The relationship between δ¹³C_{Carbon dioxide (g)}-DIC and 1/DIC is illustrated in Fig. 8a. The δ¹³C_DIC value of DR01 and DR03 was -5.6‰ and -5.8‰, respectively; these values respectively varied by -10.9‰ and -19.4‰ compared to the δ¹³C_DIC value of shallow groundwater, indicating that the DIC in Nuanquanzi thermal water mainly originated from the mantle-derived magmatic exchange and shallow groundwater due to the mixing of modified-mantle CO₂ and soil CO₂ in local areas (Koh et al., 2017). Consequently, it was presumed that the δ¹³C isotopic compositions of Nuanquanzi thermal water originated from the mixing between magmatic mantle-derived CO₂ and exchange with carbonate minerals, which was consistent with the stratigraphic lithology revealed by drilling and the existence of igneous carbonate formed from carbonate magma metasomatism during the eruption process of the Hannuoba basalt (Chen et al., 2017). The Nuanquanzi geothermal water was rich in free CO₂; the isotopic compositions of δ¹³C and δ¹⁸O indicated there exited mantle-derived magmatic CO₂ exchange in deep aquifers of the geothermal system.

Figure  8.  δ¹³C_{Carbon dioxide (g)}-DIC versus 1/DIC (a) and δ⁸⁷Sr/⁸⁶Sr versus Na⁺ (b) concentration in groundwater from Yudaokou and typical aquifers of sedimentary and volcanic geothermal fluids in China. The δ⁸⁷Sr/⁸⁶Sr values were collected from Yang et al. (2007), Ye et al. (2008), Zhai et al. (2011), Guo et al.(2020, 2009), and Ma et al. (2017).

DownLoad: Full-Size Img PowerPoint

The Sr isotopic compositions of sedimentary and volcanic geothermal fluids varied in deep aquifers such as Tengchong Rehai in Southwest China, Guanzhong Basin and Ordos Basin of Northwest China, Songliao Basin of Northeast China, and Hebei Plain and Beijing Plain of areas adjacent to Chengde, as illustrated in Fig. 8b. The ⁸⁷Sr/⁸⁶Sr ratios of the DR01 and DR03 samples were 0.707 9 and 0.707 6, respectively, which were lower than the range of values from 0.707 6 to 0.710 1 exhibited by shallow non-thermal water in the local area. Compared to the Sr isotopic compositions of typical sedimentary and volcanic geothermal systems in China (Guo et al., 2020, 2009; Ma et al., 2017; Ye et al., 2008), it is presumed that the Nuanquanzi thermal water originated from mixing with sedimentary and meteoric water.

3.2.2   Geothermal water dating

The basis of radiocarbon isotope dating is to measure the reduction of parent radioactive nucleus (¹⁴C) in a given sample. Correspondingly, the groundwater ¹⁴C dating is to determine the age of dissolved inorganic carbon (DIC) in essence (Yuan et al., 2021). The dissolution of carbonate minerals during water-rock interaction will reduce the δ¹⁴C abundance in groundwater, namely the dilution reaction. Therefore, in order to calibrate the influence of dilution effect on ¹⁴C dating of geothermal water, the calibrated age was determined after the carbon dilution factor (q) calculated by δ¹³C mixed model based on Pearson (1965).

where δ¹³C_DIC refers to the measured value of δ¹³C of groundwater dissolved inorganic carbon, δ¹³C_Soil refers to the δ¹³C value of soil CO₂ (generally close to -23‰), δ¹³C_Carb refers to the δ¹³C value of calcite (generally close to 0). After modification by the carbon dilution factor from the ¹³C isotopic compositions, the decay equation of groundwater dating can be transformed into

where t refers to the calibrated age of groundwater, λ = 12.1 × 10^-6/a, refers to the ¹⁴C decay constant, A₀ refers to the initial radiocarbon concentration of parent nucleus (pMC), generally is 100 pMC (Clark and Fritz, 1997); A_t refers to the measured radiocarbon concentration of ¹⁴C (pMC).

Relevant parameters of ¹⁴C groundwater dating were shown in Table 2. Consequently, the calibrated age of Nuanquanzi geothermal water exceeded 43.5 ka, and that of the groundwater in the upper basalt-confined aquifer was 4.05 ka, and 60 a of a typical non-thermal spring in Yudaokou. The Nuanquanzi geothermal water age was close to the upper limit of ¹⁴C dating method, owing to the long retention time of groundwater runoff. Related studies in the North China fault basin and adjacent areas demonstrate that the apparent age of Hejian deep groundwater in Baoding-Hejian profile in Hebei had exceeded the upper limit of ¹⁴C dating method (Cai et al., 2005), and the apparent age of thermal water in Xiong'an New Area were predominantly exceeded 40.0 ka (Zhu et al., 2021). The calibrated age of Hongtangsi Spring was 22 ka and that of the Tangquan Spring was 6.42 ka (Zhang, 2012). The Nuanquanzi geothermal waters in northern Chengde have experienced relatively longer residence times and larger circulation depths than those of the thermal springs in central and southern Chengde.

Table  2.  The accurate radiocarbon measurements and ¹⁴C calibrated age of water samples

No.	Sample type	Radiocarbon (pMC)	δ13C (‰)	Apparent age (a)	Dilution factor (q)	Calibrated age (a)
DR01	Thermal well	< 0.44	-5.6	> 43 500 BP	0.24	457 462.45 BP
DR03	Thermal well	< 0.44	-5.8	> 43 501 BP	0.25	473 800.40 BP
GW01	Well	60.40 ± 0.20	-19.4	4 050 BP	0.84	11 544.756 ± 38.22 BP
YS02	Spring	99.30 ± 0.40	-16.3	60 BP	0.71	5 900.09 ± 23.77 BP

 | Show Table

DownLoad: CSV

3.2.3   Elevation of the geothermal water recharge area

The δD and δ¹⁸O values decrease with the increase in meteoric water recharge elevation; thus, the recharge elevation of geothermal water can be determined according to the δD and δ¹⁸O isotopic elevation effect, which is expressed as follows (Truesdell et al., 1977)

H is the elevation of the geothermal water recharge area (m); δG is the δD or δ¹⁸O value of geothermal water (‰); δP is the δD or δ¹⁸O value (‰) of meteoric water of local area or adjacent areas; k is the isotopic height gradient (δ/100 m); h is the elevation of the water sampling point (m). The δD and δ¹⁸O values of two precipitation samples in Yudaokou were -41.07‰ and -8.17‰ and 52.75‰ and -8.01‰ (Table 1), respectively. Using the δD isotopic height gradient of -2.5‰/100 m and the δ¹⁸O value of -0.4‰/100 m, the elevation of Nuanquanzi geothermal water recharge area was calculated to be in the range of 1 800 to 2 300 m. According to the regional topographic characteristics, this result is consistent with the elevation of 2 067 m in the Daguangdingzi Peak of adjacent areas located in the northeast of the Yudaokou and Sanhanba areas in the upper reaches of the Luanhe River Basin and Liaohe River Basin at the junction of the Yinshan, Yanshan, and tail of the Greater Khingan Range Mountains.

3.3 Geothermal Reservoir Temperature

The geothermal reservoir temperature can be estimated by using empirical formulas of a geothermometer. The most widely used geothermometers are the cation temperature scale, silica temperature scale, and isotope temperature scale (Guo et al., 2020; Alçiçek et al., 2016). It is necessary to judge the water-rock interaction mineral equilibrium state of the geothermal water to ensure the reliability of the selected geothermal temperature scale before estimating the geothermal reservoir temperature using a geothermometer (Guo et al., 2016). The Na-K-Mg ternary diagram proposed by Giggenbach (1988) can be used to determine the mineral equilibrium of the geothermal system and indicate whether the geothermal water was impacted by shallow groundwater mixing (Giggenbach, 1988). As illustrated in Fig. 9, the DR02 and DR03 water samples were fully equilibrated, and DR01 was partially equilibrated close to the equilibrium line. The thermal springs in adjacent areas turned out to be immature waters, thereby indicating the disequilibrium of groundwater and host rock interaction that were more impacted by shallow groundwater mixing. Moreover, the relationship between Ca-Na-K-Mg (10C_Mg/(10C_Mg + C_Ca) and 10C_K/(10C_K + C_Ca) (Fig. 10a) were illustrated to clarify the influence of Ca²⁺ on the water-rock mineral equilibrium system (Mao et al., 2015). Nuanquanzi geothermal water samples remain fully equilibrated or close to the equilibrium line, while the other thermal springs remain in partial equilibrium.

Figure  9.  Distribution of geothermal and non-thermal water from Yudaokou and its adjacent areas in the modified Na/1 000 – K/100 – Mg^1/2 (in mg/L) ternary diagram as per Giggenbach (1988) as well as Shevenell and Goff (1995).

DownLoad: Full-Size Img PowerPoint

Figure  10.  Plots of (a) 10C_Mg/(10C_Mg + C_Ca) versus 10C_K/(10C_K + C_Ca) (Ci in mg/L), (b) log SiO₂ against the logarithm of the K²/Mg ratio on amorphous silica and quartz geothermometer equations (Giggenbach and Glovert, 1992; Giggenbach, 1988) (b), and (c) the silicon-enthalpy model (c); (d), (e), and (f) reservoir temperatures and ratio mixing of cold water defined by DR01, DR02, DR03, and shallow groundwater from Yudaokou and its adjacent areas through silicon-enthalpy function equation.

DownLoad: Full-Size Img PowerPoint

The heat dissipation was dominated by conduction-cooling of surrounding rocks during the geothermal fluid upwelling. The temperature of water samples collected from the wellhead was generally lower than that of the boiling water in the geothermal reservoir system (Giggenbach, 1988; Fournier, 1977). The heat dissipation was dominated by the adiabatic-cooling process if the geothermal fluid obtains steam-loss during upwelling (Li et al., 2015). Since the outlet water temperature of geothermal wells in Chengde was lower than 100 ℃, indicating that heat dissipation of geothermal fluid was dominated by the conduction-cooling process. So, the geothermal reservoir temperature estimation considering no steam-loss is more reliable in the Nuanquanzi area. The water-rock interaction mineral equilibrium of silicon in geothermal water was dominated by quartz, chalcedony, and amorphous SiO₂. The reservoir temperatures and the mineral equilibrium of silicon were defined by plotting the logarithm of silica concentration versus K²/Mg(lg[C(SiO₂)] – lg[(C(K))²/C(Mg)]) (Mao et al., 2015; Giggenbach, 1988). As illustrated in Fig. 10b, the mineral equilibrium of silicon in the Nuanquanzi geothermal water was dominated by quartz. The geothermal reservoir temperature calculated by quartz geothermometer was more applicable than other silica geothermometers.

The geothermal reservoir temperature estimated by chalcedony and K-Mg geothermometer was lower than the actual measured temperature in DR01 and DR02 at the well bottom (61.6 ℃) (Table 3). The geothermal reservoir temperature calculated by Na-K geothermometer was higher than the K-Mg geothermometer (except DR01). The mineral equilibrium of K-Mg was constantly broken and rebalanced by the conduction-cooling or cold water-mixing processes during geothermal fluid upwelling, leading that the calculation results of K-Mg geothermometer tend close to the heat-exchange temperature in the shallow thermal reservoir or even the temperature of water samples collected from the wellhead (Guo et al., 2016). In summary, the Na-K geothermometer is less affected by the rebalancing of the mineral equilibrium system and the cold water-mixing process during geothermal fluid upwelling, and the equilibrium of SiO₂ in thermal water was dominated by quartz. The geothermal reservoir temperature calculated by Na-K geothermometer (DR02), and by quartz geothermometer (without steam separation or mixing) (DR01) was relatively more reliable. Accordingly, the reservoir temperature of the Nuanquanzi geothermal system was determined to be 73.39–92.87 ℃, matching well to the geothermal reservoir (80–100 ℃) in the Kangbao-Waichang deep fault zone (Zhang, 2012).

Table  3.  Temperature calculations performed via the cation and silica geothermometer of Nuanquanzi geothermal water

Geothermometer (℃)		DR01	DR02	DR03
TNa-K		68.88	92.87	83.42
TK-Mg		71.74	58.11	53.94
TNa-K-Ca		85.9	75.16	71.77
TSiO2	T1	73.84	83.39	63.65
	T2	73.39	82.87	63.44
	T3	77.75	86.08	68.95
	T4	41.72	51.71	31.37
$ {T}_{\mathrm{N}\mathrm{a}-\mathrm{K}}=\frac{1\ 390}{\mathrm{l}\mathrm{o}\mathrm{g}\left(\left[\mathrm{N}\mathrm{a}\right]/\left[\mathrm{K}\right]\right)+1.75}-273.15 $ (Giggenbach, 1988), where [Na] and [K] represent the concentrations of Na and K in mg/L, respectively. $ {T}_{\mathrm{K}-\mathrm{M}\mathrm{g}}=\frac{4\ 410}{\mathrm{l}\mathrm{o}\mathrm{g}\left(\left[\mathrm{K}\right]/\sqrt{\left[\mathrm{M}\mathrm{g}\right]}\right)+14.00}-273.15$ (Giggenbach, 1988), where [K] and [Mg] represent the concentrations of Na and K in mg/L, respectively. $ $$\begin{array}{l}{T}_{\mathrm{N}\mathrm{a}-\mathrm{K}-\mathrm{C}\mathrm{a}}=4\ 410/\left\{\begin{array}{l}\mathrm{l}\mathrm{o}\mathrm{g}\left(\left[\mathrm{N}\mathrm{a}\right]/\left[\mathrm{K}\right]\right)+\\ \beta \left[\mathrm{l}\mathrm{o}\mathrm{g}\left(\sqrt{\left[\mathrm{C}\mathrm{a}\right]}/\left[\mathrm{K}\right]\right)+2.06\right]+2.47\end{array}$$ \right\}\\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ -273.15\end{array} $ (Fournier and Truesdell, 1973), where [Ca], [Na], and [K] represent the concentrations of Na and K in mg/L, respectively. β = 4/3 when T < 100 ℃ and β = 1/3 when T > 100 ℃. T1 represents the temperature estimated by the quartz geothermometer without steam separation or mixing (Truesdell, 1977): T1 = -42.198 + 0.288 31[SiO2] – 3.668 6 × 10-4[SiO2]2 + 3.166 5 ×10-7[SiO2]3 + 77.034 1lg[SiO2], where [SiO2] is the aqueous SiO2 in mg/L. T2 represents the temperature estimated by the quartz geothermometer without steam loss (Truesdell, 1977): T2 = 1 309/(5.19 - log[SiO2]) - 273.15, where [SiO2] is the aqueous SiO2 in mg/L. T3 represents the temperature estimated by the quartz geothermometer of silica-quartz conductive cooling (Fournier, 1977): T3 = 1 522/(5.75 - log[SiO2]) - 273.15, where [SiO2] is the aqueous SiO2 in mg/L. T4 represents the temperature estimated by the chalcedony geothermometer (Fournier, 1977): T4 = 1 032/(4.69 - log[SiO2]) - 273.15, where [SiO2] is the aqueous SiO2 in mg/L.

 | Show Table

DownLoad: CSV

3.4 Impact of Geothermal and Non-thermal Water Mixing

According to the hypothesis of silicon-enthalpy model of geothermal fluid, the soluble SiO₂ in the deep geothermal fluid is generally saturated, and the soluble SiO₂ concentration decreases with the upwelling of geothermal fluid owing to the cold water mixing or precipitation during water-rock interactions (Rybach and Muffler, 1987). The Nuanquanzi thermal water was demonstrated to originate from a relatively open thermal reservoir environment and affected by cold-water mixing during the geothermal water convection in the fault zone. Therefore, the following silicon-enthalpy function equation can be constructed to estimate the proportion of cold-water mixing (Fournier and Truesdell, 1973)

where S_c is the enthalpy of shallow cold water, S_h is the initial enthalpy of deep geothermal water, S_S is the final enthalpy of geothermal water, SiO_2c is the soluble SiO₂ concentration of shallow cold water, SiO_2h is the soluble SiO₂ concentration of geothermal water, and SiO_2S is the final soluble SiO₂ concentration of geothermal water. X is the proportion of cold-water mixing (Fournier, 1977). In addition, the proportion of cold-water mixing and initial geothermal reservoir temperature can be estimated by SiO₂-enthalpy diagram (Fig. 10c), of which the applicable conditions were that the quartz was the dominated mineral controlling the concentration of soluble SiO₂ in geothermal fluid, and secondly, there shall be no dissolution and precipitation of SiO₂ heat dissipation during geothermal fluid upwelling. As indicated by Figs. 10a, 10b, the soluble SiO₂ equilibrium system of DR01 sample dominated by quartz, and that of DR02, and DR02 sample that full equilibrated demonstrated to be affected only by cold-water mixing without steam-loss. It is suitable to use silicon-enthalpy model to estimate the mixing proportion of cold water.

Taking the average soluble SiO₂ concentration of 13.42 mg/L as the SiO_2c value and the average temperature of 6.9 ℃ as the parameter associated with enthalpy S_c of shallow groundwater around Nuanquanzi (Sun et al., 2020), the intersection points of the quartz curve of steam-loss modification line of the Nuanquanzi geothermal water indicated a reservoir temperature of 120 to 187.5 ℃ (Fig. 10c).

The proportion of cold-water mixing in DR01, DR02, and DR03 sample were 90.8%, 88.3%, and 92.2%, while the reservoir temperatures defined by DR01, DR02, and DR03 were 170, 168, and 167 ℃, respectively (Figs. 10d, 10e, 10f). In summary, the proportion of cold-water mixing with the initial deep geothermal water of the Nuanquanzi geothermal system was approximately 88.3% to 92.2%, much larger than that of deep geothermal water (42% to 67%) at the middle of the North China fault depression basin (Zhu et al., 2021).

The initial deep geothermal reservoir temperature estimated by SiO₂-enthalpy diagram (167 to 170 ℃) was much higher than the cation and silica geothermometer (73.39 to 92.87 ℃). It is related to the conduction-cooling process during the parent geothermal fluid upwelling (Li et al., 2015). SiO₂ dissolution or precipitation and heat loss occurred during the mixing of the geothermal fluid and shallow cold water; namely, there was steam loss or heat dissipation to the surrounding rocks. On the other hand, the SiO₂-enthalpy model was established according to the ideal mixed-mode considering only single factor such as the mixing with single strand cold water. The estimation results based on the ideal state were quite different from the actual complex situation, such as the mixing multiple small streams of cold water (Yan et al., 2019). The formation of geothermal anomaly in the mountain areas of North China is primarily related to the heat source from the upper mantle. In addition, the heat from the upper mantle, and the decay heat from radioactive elements migrate upward to the shallow thermal reservoir with geothermal fluid as the heat carrier through the conduction channel formed by deep and large fault structures. The infiltrated precipitation obtained thermal energy from the deep crustal heat source, then upwelling to the shallow part of the fault zone where the cold water mixed with the geothermal fluid after deep circulation (Liu et al., 2020).

3.5 Geothermal Water Circulation Depth

The Nuanquanzi thermal reservoir was a medium-low temperature geothermal system formed by deep circulation of heating conduction during meteoric water recharging. Thus, the circulation depth of geothermal water can be estimated according to the geothermal gradient measurement, and the calculation formula can be expressed as follows

where H is the geothermal water circulation depth (m), T_r is the thermal reservoir temperature (℃), T₀ is the temperature of the constant temperature zone, t is the ground temperature gradient (℃/100 m), and h₀ is the depth of the continuous temperature zone. As measured and shown in Fig. 2, the average geothermal gradient of Nuanquanzi DW01 was 2.87 ℃/100 m, and that of DW02 was 2.81 ℃/100 m. Taking the annual average temperature in the Yudaokou area of 4 ℃ as the T₀ value and the depth of the constant temperature zone of approximately 15 m as the h₀ value, the geothermal water circulation depth was estimated to range from 2 400 to 3 200 m; this range was close to that of the geothermal water circulation depth around the Kangbao-Waichang deep fault zone (2 274 to 3 645 m).

3.6 Geotectonic Genesis: Evidence from Integrated Geophysical Exploration

3.6.1   Distribution characteristics of soil radon

The hydrochemical field, temperature field, and structural conditions of the geothermal system are conducive to the precipitation and migration of radon gas in groundwater (Roba et al., 2012). The fault-exposed area is usually accompanied by radon anomalies, and the soil radon measurement is an effective technical means to detect faults and indicate the genesis of geothermal water (Tabar et al., 2013).

The curve of radon amplitude variation was obtained along with a contour map of the spatial distribution of the soil radon content in Fig. 11. The average value of soil radon in the 95% confidence interval of the survey area was 1 935.5 Bq/m³. The height anomaly single peak or multi-peak appeared in the radon measurement curve in the areas of the spring outcrop zone of Nuanquanzi and Hounuanquanzi, as well as in the Xiaoluan and Ruyi River coastal areas. The high radon concentration points of Nuanquanzi extend from northwest to southeast in a strip-shaped shape and are distributed in a sheet shape around Hounuanquanzi. There are high abnormal areas distributed in the Nuanquanzi and Hounuanquanzi areas around geothermal wells, as shown in the contour map of radon gas concentration (Fig. 11c). The concentrated range of the Nuanquanzi-Huangtushan area was distributed in a belt-shaped trap state, spreading from northwest to southeast in a parallel direction to the Kangbao-Weichang deep fault; this result was consistent with the strike of the NW-SE normal fault of the Yudaokou Ranch area. The other abnormal area of soil radon was distributed along the Xiaoluan River from Huangtushan to Bashang Village in the SW to NE direction, and a secondary anomalous trap zone accompanied by the Xiaoluan River abnormal area was located in the northeast of Huangtushan. Furthermore, these extension directions of abnormal zones were approximately parallel to the northeast upward Huangqi-Wulonggou deep fault in Chengde (F5 in Fig. 1), and consistent with the regional secondary Xiaoluan River fault belt in Yudaokou. Hounuanquanzi turned out to be the concentration center of soil radon, where the high anomaly zone was distributed in a sheet shape. Molten asphalt was found at a depth of 23.5 m of the drill in Hounuanquanzi (Fig. 11), which was associated with the distribution of soil radon; it can be presumed that the Nuanquanzi geothermal field was located at the intersection of regional northeast and northwest faults. The anomaly of soil radon in the northwest of the area was also distributed as a band shape; however, it was not trapped, with its extension direction being parallel to the Shanghuangqi-Wulonggou deep fault presumably due to inherited secondary concealed faults in the area.

Figure  11.  Molten asphalt found at the drill in Hounuanquanzi (a), variation curve of soil radon concentration in survey lines (b), and contour map of the spatial distribution of soil radon and inferred faults in Yudaokou area (c).

DownLoad: Full-Size Img PowerPoint

3.6.2   Interpretation of CSAMT detection

The anomaly of soil radon measurement is generally considered to be directly related to the distribution of the structure on the space plane. However, the vertical characteristics of structural faults, such as extension depth and occurrence, cannot be obtained effectively. The CSAMT is characterized by considerable detection depth, high lateral resolution, and less affected by topography. Thus, it can detect the favorable structures for geothermal conduction and reservoir and determine the width and downward extension depth of the anomaly zone effectively (Soengkono et al., 2013).

The integration atlas of the CSAMT apparent resistivity inversion profiles, which presumed the geological interpretation and the radon amplitude distribution curve, is illustrated in Fig. 12 based on drilling, geophysical interpretation, and soil radon measurement. R_nB is the soil radon background value determined by the average radon concentration in the 95% confidence interval of each section. RnF is the anomaly threshold of the radon content calculated using the sum of RnB and the square of standard deviation. The R_nB and RnF values of PM01 were 1 235 and 2 023 Bq/m³, while those of PM02 were 1 495 and 2 398 Bq/m³, respectively. There were two regions where the radon concentration exceeded the high abnormal threshold (RnF) on the radon amplitude distribution curve: from 0–100 m and at approximately 1 400 m of the CSAMT detection line of PM01 (Fig. 12a); with the maximum value of 2 859 Bq/m³. The radon concentration between 1 000 and 1 500 m of the PM01 line was higher than the background value, showing a multi-peak anomaly. The apparent resistivity at 1 350 m was high on both sides and low in the middle; there were rising springs exposed on the surface, indicating that faults were distributed under the ground. The apparent resistance values of the underlying strata show a sharp discrepancy in the 900– 1 100 m part of the PM01 line, where the value in the northwest of the profile exceeds 900 Ω·m; however, that in the northeast is lower than 50 Ω·m. The depth of the apparent resistance differentiation zone was approximately 300 m; the high-resistivity anomaly zone in the lower basement was uplifted here, indicating that there was a relatively deep fault that had developed underground. The resistivity of dense basalt was higher than that of the vesicular basalt; therefore, the discrepancy in apparent resistivity in the shallow part of the profile was mainly due to the distribution of different basalt lithofacies caused by fault dislocation. There was a low-resistivity zone between two high-resistivity zones from 1 850 to 2 050 m of the PM01 line that was shallower than 100 m; this zone was presumed to be a hydraulic fracture zone with low resistivity that passes through the dense olivine basalt zone with high resistivity based on hydrogeological drilling of drinking water in the local area.

Figure  12.  Integration atlas of soil radon amplitude distribution curve, CSAMT apparent resistivity inversion (b), and presumed geological interpretation section of PM01 and PM02 geophysical lines in Yudaokou area.

DownLoad: Full-Size Img PowerPoint

Radon concentrations exceeded the high abnormal threshold (RnF) on the radon amplitude distribution curve at two locations near the Xiaoluan River; that is, at 2 400 and 2 850 m of the CSAMT detection line of PM02 approximately (Fig. 12b), with peak values of 6 693 and 4 063 Bq/m³, respectively. There was a high-resistivity zone with an extended depth of approximately 800 m at 2 600 m of the PM02 line. The apparent resistance values of adjacent areas show a sharp discrepancy close to 10 Ω·m. The low-resistivity zone in the shallow part of the profile turned out to be a permeable and water-free Quaternary aeolian sand layer. There was a good correspondence between the high radon anomaly and low-resistivity zone, which were presumed to be two faults distributed on both sides of the dense basalt. At the part of the PM02 line from 0 to 300 m, there was a clear boundary between the high and low resistivity zones with the low-resistivity belt downcutting steeply to the northwest, which is presumed to be a fault. Based on the two geothermal well drilling depths of 1 300 and 2 000 m and the measurement of rock physical parameters, the low-resistivity zones turned out to be shallow loosely deposited Quaternary aquifers and vesicular basalt aquifers with secondary caves and paleo-weathering crusts in the interval of multi-stage magma eruption; the high-resistivity zones were dense basalt aquicludes formed by upwelling of mantle magma. The underlying bedrock of the mat-shaped distribution basalt was thick laminar rhyolitic tuffaceous lava, rhyolite, and andesite with uniform thickness. The stratum of the high-resistance zone of the basement below 1 000 m was rhyolite and rhyolitic tuff.

In general, the application of CSAMT interpretation and radon gas measurement provided strong evidence for detecting hidden structural faults in the study area while elucidating its geotectonic genesis. The favorable stratigraphic pattern of the graben basin for the geothermal reservoir was formed by the multiple faults caused due to stratum depression in the southwest and local uplift in the northeast of the area. The dense thick-bedded Cretaceous tuff lava, Upper Jurassic tuff, and sedimentary mudstone with poor water permeability and low thermal conductivity play a key role in the heat preservation for geothermal reservoirs, which are considered to be the caprock of the geothermal system. The Jurassic sandstone with fairly good water abundance was considered the geothermal reservoir, while the multi-stage structural fracture zone functioned as the main upward channel of geothermal water.

3.7 Geothermal Origin Model

The Nuanquanzi geothermal reservoir occurs in the Yanshan uplift belt in the North China intracontinental orogenic belt (Chen et al., 1990; Chen, 1988). The Nuanquanzi geothermal system was indicated to be a medium-low temperature convection type, and the thermal reservoir was found to be the fault semi-trap type. The origin of geothermal water was dominated by the mantle subplume and mantle branch structure, while heat stems from the rapidly emplaced magmatic rocks and densely developed ductile shear or fault zones (Chen et al., 1990; Chen, 1988). Magma upwelling was accompanied by mantle upwelling in the Yanshan belt; thus, it was a critical heat source of the geothermal system of the Yanshan uplift belt in the deep geothermal water at the middle of the North China fault depression basin that exhibits convection circulation along the ductile shear zone under the compression caused by the destruction of the North China Craton. Therefore, the terrestrial heat flow value and geothermal reservoir temperature in the Yanshan uplift belt were relatively lower than those in the Central North China fault depression basin (Wang et al., 2017). The thermal reservoir is a pore and fissure aquifer of the Jurassic volcanic sedimentary rock association that consists of tuffaceous sandstone and glutenite. Moreover, the caprock of the geothermal system was mat-shaped, densely distributed, thick-bedded Cretaceous tuff lava, and Upper Jurassic tuff originating from magma overflow. The E-W faults and NE-NNE faults formed by the movements of Yanshan, Himalayas, and neotectonics intersect in the Nuanquanzi area, creating a favorable channel for water circulation and heat conduction of geothermal fluid. Several local second-order fault basins were formed after the multi-stage fault tectonic movement, with some strata being raised and depressed. The heat flow redistributes and concentrates to the sandstone bulge area, while the geothermal anomaly was formed during the upward conduction of the deep heat source; this is because the thermal conductivity of the sandstone layer was higher than that of the upper layer. The boundary faults of the sandstone bulge area also serve as the upwelling channels of deep geothermal water, providing a convective driving force for the geothermal system. Based on conventional groundwater geochemistry and the isotope database, the regional groundwater flow direction is generally northeast to southwest, the Nuanquanzi geothermal water is of meteoric origin, and the recharge area is the Daguangdingzi peak with an elevation of 2 067 m at the junction of the Yinshan, Yanshan, and the tail of the Greater Khingan Range Mountains. The residence time of geothermal water was approximately 43.5 ka, and the temperature of recharge water was calculated to be -1.06 to 2.32 ℃. Moreover, the geothermal reservoir temperature geothermal system ranges from 73.39 to 92.87 ℃ after atmospheric precipitation is recharged from the northeast, heated, and mixed by high-temperature primary geothermal water and steam during forced-convection circulation at a depth of 2 400 to 3 200 m. According to the silicon-enthalpy model and geothermometer, the pre-mixing temperature of the geothermal reservoir in the deep ductile shear zone and fault zone ranged from 167 to 170 ℃, and the proportion of cold-water mixing during circulation was approximately 88.3% to 92.2%. The presumptive geothermal genetic model is shown in Fig. 13.

Figure  13.  Presumptive geothermal genetic model of Nuanquanzi geothermal water in Yanshan uplift belt.

DownLoad: Full-Size Img PowerPoint

4. CONCLUSIONS

The Nuanquanzi geothermal system is a medium-low temperature convection type, and the thermal reservoir was demonstrated to be the fault semi-trap type with a geothermal gradient of 2.81–2.87 ℃/100 m. The hydrochemical type of thermal water is primarily HCO₃-Na, and rich in soluble SiO₂, F^- and Cl^-. The ionic characteristics of the geothermal water inherit the mineral characteristics of the surrounding rock in the fault zone.

The geothermal water primarily originated from meteoric water but affected by the mixing of endogenous sedimentary water. The Daguangdingzi peak with an elevation of 2 067 m was determined to be the Nuanquanzi geothermal system's recharge area, and maximum circulation depth of geothermal water was estimated to be 2 400–3 200 m. The calibrated ¹⁴C age of the Nuanquanzi geothermal water exceeded 43.5 ka, significantly higher than that of the thermal springs in central and southern Chengde. The geothermal water in northern Chengde obtained relatively longer residence time and considerable circulation depth than those of the thermal springs in Central and South Chengde.

The reservoir temperature calculated by Na-K and quartz geothermometer of the Nuanquanzi geothermal system was determined to be 73.39–92.87 ℃, which is lower than the pre-mixing temperature of 167 to 170 ℃ based on the silicon-enthalpy model. The conduction-cooling processes, namely steam loss, heat dissipation to surrounding rock, and shallow cold-water mixing processes, occurred during the parent geothermal fluid ascent to the surface. The proportion of cold-water mixing with the initial deep geothermal water of the Nuanquanzi geothermal system was approximately 88.3% to 92.2% during circulation.

The comprehensive geophysical application of soil radon measurement and CSAMT interpretation provide solid evidence for identifying structural faults in geothermal reservoirs. The high radon concentration anomaly area matched well to the low apparent resistivity and the distribution of insidious faults. The Nuanquanzi geothermal field was developed at the intersection of regional northeast and northwest faults. The multiple faults formed in different orogenic periods created a favorable stratigraphic pattern of the graben basin for the geothermal reservoir. Geothermal resources in the geothermal field of the Yanshan uplift belt exhibit high potential for exploitation and utilization, owing to the heat convection circulation along the ductile shear zone under the compression caused by magma upwelling and regional relatively thick crust caused by the thinning and destruction of the lithospheric mantle root of the North China Craton.