Table 2 shows that the water chemistry of geothermal water in the Woka geothermal area is dominated by Na⁺, Ca²⁺, and K⁺ in cations and SO₄ ²⁻ and Cl⁻ in anions, followed by HCO₃ ⁻. The pH values and total dissolved solid concentrations are 7.95–8.25 and 285.90–440.51 mg/L, respectively. Geothermal waters belong to medium–high alkaline water. The hydrochemistry types are mainly SO₄-Na type, Cl-SO₄-Na type, and HCO₃-SO₄-Na type.

The hydrochemical characteristics of geothermal water and bedrock fracture cold water in the fault rift basin are very different (Figure 2). The geothermal waters are enriched in H₂S but poor in CO₂. Their hydrochemical type is influenced by H₂S, and there are more complex hydrochemical types such as SO₄ type, SO₄-Cl type, and HCO₃-SO₄ type, consistent with the Riduo geothermal spring and Gudui geothermal field in the fault rift zone. The hydrochemical type of the Luobusa geothermal spring is different from that of geothermal waters in the Woka rift, indicating that they do not belong to the same geothermal system.

As shown in the Schoeller diagram (Figure 3), all ions in the geothermal water have regular variations, revealing that the source and formation pattern of geothermal water in the basin have similar characteristics. In addition, the geothermal waters have high SO₄ ²⁻ contents, consistent with the strong sulfur smell of the geothermal water outcrop, as typified by the Woka geothermal spring. The source of SO₄ ²⁻ in geothermal water is mainly from geothermal water dissolution filtration of sulfate rocks, and a small part is H₂S carried by deep fluids. Therefore, the higher SO₄ ²⁻ contents in the geothermal water of Woka mainly come from the strong dissolution filtration of volcanic clastic rocks such as andesitic and rhyolitic lava dominated by the Yeba Formation strata. The saturation indices of geothermal water calculated using Phreeqc software are mostly lower than zero. The results show that most minerals, except for chalcedony and quartz, are unsaturated (Table 3), reflecting the short runoff pathway and weak water–rock reaction.

Four geothermal water samples were taken for δ¹³C analysis. The δ¹³C values of geothermal spring water are −6.49‰ to −3.45‰: the δ¹³C value of Kanai geothermal spring is −6.1‰, the δ¹³C value of Woka geothermal spring is −6.49‰, the δ¹³C value of Misbah geothermal spring is −5.41‰, and the δ¹³C value of Lobsa geothermal spring is −3.45‰ (Figure 4). Previous studies showed that the δ¹³C value of upper mantle material source is −8‰ to −4‰, the δ¹³C value of carbonate metamorphic source is −2‰ to +2‰, and the δ¹³C value of carbonate metamorphic source is −2‰ to +2‰. Therefore, it is presumed that mantle-sourced CO₂ from deep faults was involved in the formation of geothermal waters in the study area.

Four geothermal water samples were selected for the analysis of hydrogen and oxygen isotopes. The results show that the δ²H values range from −150.8‰ to −156.3‰ and the δ¹⁸O values range from −19.46‰ to −19.86‰. In Figure 5, all samples are distributed near the global atmospheric precipitation equation line, indicating that the source is recharged by atmospheric precipitation and less infiltration of snow-melting water. The deuterium excess parameter (d) is similar between the geothermal spring and the cold spring, indicating that there is no obvious oxygen drift in the geothermal water. The result reveals the characteristics of atmospheric precipitation recharge and shallow circulation in the geothermal water system of the Woka rift. The deuterium excess parameter d of the Luobusa geothermal spring is −33.82, which shows significant oxygen drift, further supporting that it belongs to a different geothermal system from the Woka graben, with deep circulation recharge characteristics.

The isotopic composition of atmospheric precipitation has an elevation effect, and the recharge elevation of geothermal springs can be calculated using δ¹⁸O isotopic values based on this feature (Zhang et al., 2018). The calculation equation is as follows: H = δ G − δ P K × 100 + h (1) where H is the elevation of the recharge area, m; h is the elevation of the sampling point, m; δ_G is the δ¹⁸O value of the geothermal spring; δ_P is the δ¹⁸O value of atmospheric precipitation; and K is the δ¹⁸O elevation gradient of atmospheric precipitation (δ/100 m). The gradient value of δ¹⁸O is determined as follows: four typical cold springs were selected along the Zengjiuqu from upstream to downstream, and the δ¹⁸O was obtained by fitting the δ¹⁸O to the outcrop elevation of the spring. The curve is δ¹⁸O = −0.0037*H-107.98, R ² = 0.9473, and the height gradient obtained from the fit is 0.37‰/100 m, which is in good agreement with the height gradient of 0.31‰/100 m in Tibetan area. The elevation reference value is determined as follows: The lake water is used as the elevation reference point, and its δ¹⁸O = −15.0‰. The calculation results of the recharge elevation of each geothermal spring are shown in Table 4, and the results show that the recharge elevation of the geothermal water system in the Woka graben basin is 5193–5247 m, which is consistent with the elevation of the alpine basin where the regionally distributed snow mountains are located.

The temperature of geothermal springs, as natural outcrops of geothermal water systems, is often low because of the mixing of shallow groundwater or surface water and cannot actually represent the temperature of the geothermal reservoir (Zhang, Xu, 2018; Li et al., 2020; Chang et al., 2021). Therefore, quantitative geochemical geothermometers such as SiO₂ and cation geothermometers are usually used to calculate the reservoir temperature (Fournier, 1977; Truesdell and Fournier, 1977; Fournier, 1979). As seen in Figure 6, the Kanai geothermal springs and the Woka geothermal springs all fall in the partially equilibrated zone or mixed water zone, and the Cuoba geothermal springs and the Luobosha geothermal springs fall in the immature water zone, indicating that the water–rock interaction in the geothermal water system of the Woka fault rift basin has not reached to complete equilibrium, or the geothermal water is mixed with a large proportion of cold water in a nonequilibrium state. Hence, the cation geothermometer has limitations in estimating the temperature of the reservoir. In this study, the SiO₂ geothermometer is selected for the calculation of the geothermal reservoir. The geothermal reservoir temperatures range from 95.22°C to 121.66°C (no steam loss) and 96.85°C–119.57°C (maximum steam loss), respectively, measured using the SiO₂ geothermometer. From the Na-K-Mg equilibrium diagram (Figure 6), it can be seen that the four outcrops of the Woka geothermal spring all fall between the 120°C and 200°C isotherms. In addition, the estimated reservoir temperatures based on the SiO₂ geothermometer are all lower than this range, indicating that the geothermal water in the basin is mixed by the shallow fractured cold water or surface water in the process of gushing out of the surface.

The above analysis shows that the geothermal springs in the Woka basin have the characteristics of mixing with shallow cold water, and a mixing model can be established to estimate the initial geothermal reservoir temperature in the deep position. The mixing model equation is as follows: S c X 1 + S h ( 1 − X 1 ) = S s (2) ρ c SiO 2 X 2 + ρ h SiO 2 X 2 ( 1 − X 2 ) = ρ S SiO 2 (3) where S_c is the enthalpy of cold water (J/g); S_s is the final enthalpy of spring water (J/g, the enthalpy of saturated water below 100°C is equal to the number of the temperature of water in degree Celsius; above 100°C, the relationship between temperature and enthalpy of saturated water can be found in Table 5); S_h is the initial enthalpy of geothermal water (J/g); ρcSiO₂ is the mass concentration of SiO₂ of cold water (mg/L); ρsSiO₂ is the mass concentration of SiO₂ of spring water (mg/L); ρhSiO₂ is the initial mass concentration of SiO₂ of geothermal water (mg/L); and X is the mixing ratio of cold water.

According to the above equation, the cold water temperature is taken as the local average annual temperature of 8.2°C in Woka, and the cold water SiO₂ content is taken as the average value of cold springs in the basin. The initial temperature of geothermal water is assumed to be equal to 75°C–300°C, and the corresponding enthalpy values can be found in Table 6. The temperature at the mouth of the spring and the SiO₂ content were based on the measured values. The enthalpy and the mass concentration of SiO₂ at each temperature are substituted into Eqs 1, 2, respectively, to find the X₁ and X₂ values at different temperatures and plot the curve of X and geothermal water temperature at different temperatures (Figure 6), and the intersection point of X₁ and X₂ curves is the calculated reservoir temperature value.

Figure 7 shows that the reservoir temperature of Kanai geothermal spring is 200°C, and the proportion of cold water mixing is 77%; the reservoir temperature of Woka geothermal spring is 175°C, and the proportion of cold water mixing is 73%; the reservoir temperature of Amaka geothermal spring is 175°C, and the proportion of cold water mixing is 83%; thus, it can be seen that the underground geothermal water system of Woka rift basin is a high-temperature geothermal water system, but it is strongly affected by the mixing of cold water on the surface. The estimated reservoir temperature based on the mixing model is basically consistent with the reservoir temperature range on the Na-K-Mg equilibrium diagram, which can reflect the reservoir temperature value of the subsurface geothermal water system in the Woka graben basin more realistically.

Because of the continuous and strong plate collision activities of the Eurasian plate, intensive and frequent structural deformation has been formed on the Tibetan Plateau. It creates a deep melt-type magma source area in the lower crust and a local low-velocity melt layer in the middle and forms shallow intrusive magma capsules and local banded melt bodies at different depths within a certain depth in the upper crust, providing the necessary heat source for the formation of the underground geothermal water system on the Tibetan Plateau. Therefore, the Tibetan Plateau geothermal resources are mostly uplifted mountain convection type. The conceptual model of the subsurface geothermal water system within the Woka graben basin was established by synthesizing the research results of many geothermal fields in the Tibetan region, such as Yangpaijing and Nimu-Naqu geothermal fields (Figure 8).

The atmospheric precipitation and alpine ice and snow melt water on both sides of the Woka graben basin are the main recharge sources of geothermal water, and the cold water receiving the recharge keeps seeping down along the fractures developed by the tectonic activity. In addition, cold water is heated by the local strip melt at a certain depth to provide a heat source and thus forms geothermal water involved with deep gas components. Then, geothermal water begins to flow upward, controlled by the fault-controlled bedrock fracture type reservoir. During the ascent of geothermal water along the river valley or the basin boundary fractures, its temperature decreases because of the mixing of the shallow cold water. In Zengjiuqu River valley, geothermal water enters into the pore aquifer of the fourth system loose rock type with good permeability along the bedrock fracture reservoir. In addition, the dense sand layer or geothermal water cementation layer is used as the cover layer, forming the Quaternary pore type shallow reservoir, which is mixed with the surface water during the rising of geothermal water and forms a warm water swamp along the riverbed.

The Woka graben is adjacent to Mochukongka Qu in the north and reaches the valley of Yarlung Tsangpo River in the south, which is an essential place for the Sichuan–Tibet railway line. Because of the construction of the Lhasa–Linzhi railway, many major projects such as hydropower and road construction are further planned in the area. According to the above analysis, the geological structure of Woka graben is complex, with active fracture development and strong hydrothermal activity, and its deep reservoir temperature can reach more than 120°C. Hence, it belongs to the high- and medium-temperature geothermal system. Engineering construction will be affected by the enrichment of geothermal resources.

As for the engineering thermal damage, the water temperature of the surface outcrop in the Woka rift valley is 36.25°C–57.77°C, and the calculated reservoir temperature is 97.86°C–125.56°C. The water temperature in the deep or good geothermal channels will be even higher. After the project exposure, there will be a large flow of high-temperature geothermal water caused by the surge water disaster; the sulfate content of geothermal water in the graben is 94.4–147 mg/L, and toxic and harmful gases such as H₂S will be escaped after the project exposure. Therefore, the construction of the project in the geothermal development area of the rift valley will pose a great threat to the safety of construction workers because of the high-temperature geothermal water and harmful gases, leading to a significant increase in project costs.

As for the corrosiveness of the project, the pH value of the geothermal water in the rift valley is 6.71–8.25, the HCO₃ ⁻ content is 31–214 mg/L, the Mg²⁺ content is 0.026–7.85 mg/L, and the SO₄ ²⁻ content is 94.4–147 mg/L, which is not corrosive to the project. However, the Cl⁻ ion content of the Luobusa geothermal spring is 4060 mg/L, which is more than 1000 mg/L. It is a corrosive hazard to the tunneling equipment as well as the tunneling facilities, which must be considered in tunnel construction. Therefore, for major construction projects in the Woka graben, a detailed investigation should be conducted to avoid the geothermal water channel or to divert the geothermal water and reasonably arrange the construction sequence to overcome the problem.