Abstract:Water management has increasingly become a hotspot in shale gas supply chains. This study develops a comprehensive modeling framework for the mixed noise-driven shale gas-water supply chains, which is integrated with techniques of system dynamics model and two-stage stochastic risk-aversion programming. White and colored noises are used for addressing stochastic characterization of shale gas productivity. Regional water resources carrying capacity is reflected based on the system dynamics model. The developed model cannot only effectively address stochastic parameters in the objective and constraints, but also offer a linkage between the pre-regulated policies and corresponding economic implications raised from improper policies. A Marcellus-based case in Pennsylvania is then performed to validate the applicability of the developed model. Results reveal that shale gas production curve is more sensitive to colored and mixed noises than white noise. High flowback and recycle rates would increase carrying capacity of shale wells by 1.8% in 2023. An increase in weighting factor (e.g., from 0 to 30) would induce a rise in system cost (e.g., from $1.22×1011 to $1.34×1011) but a reduction in conditional value-at-risk value (e.g., from $6.83×1010 to $3.83×1010). Weighting factor thus can be served as an indicator to show how much the decision makers' attention to system risk.
environmental sciences,green & sustainable science & technology,engineering, environmental
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to optimize water resource management in the shale gas - water supply chain. In particular, under the drive of mixed noise, how to improve the efficiency of this process through system dynamics analysis and risk - averse planning. Specifically, the paper focuses on the following aspects:
1. **Assessment of water resource carrying capacity**: Evaluate the regional water resource carrying capacity in shale gas development through a system dynamics model to ensure the rational use of water resources while meeting the water demands of industry and residents.
2. **Impact of mixed noise**: Research the influence of white noise and colored noise on the shale gas production curve, especially how these noises affect the Estimated Ultimate Recovery (EUR) value of shale gas. This helps to predict the production dynamics of shale gas more accurately.
3. **Risk - averse planning**: Optimize the decision - making process of the shale gas supply chain through a Two - Stage Stochastic Risk - averse Programming (TSRP) model combined with the Conditional Value - at - Risk (CVaR) method. This model can provide more robust economic and environmental decisions on the basis of considering uncertainty.
4. **Integrated modeling framework**: Develop an integrated modeling framework that combines the system dynamics model, the two - stage stochastic risk - averse programming, and the noise - enhanced Arps method to deal with multiple uncertainties in the shale gas - water supply chain.
### Main contributions of the paper
1. **Innovatively combined multiple techniques**: Combine system dynamics analysis, two - stage stochastic risk - averse programming, and the noise - enhanced Arps method to form an integrated modeling framework, which improves the practicality and accuracy of the model.
2. **Detailed assessment of water resource carrying capacity**: Through the system dynamics model, the regional water resource carrying capacity in shale gas development is evaluated in detail, providing a scientific basis for water resource management.
3. **Analysis of the impact of mixed noise**: The influence of white noise and colored noise on the shale gas production curve is studied, revealing the sensitivity of different types of noise to the production dynamics of shale gas.
4. **Optimized supply chain management**: Through the two - stage stochastic risk - averse programming model, the management strategy of the shale gas supply chain is optimized, improving the economic and environmental sustainability of the system.
### Formula summary
- **System dynamics model**:
- The system dynamics model mainly involves state variables, rate variables, and feedback control.
- The specific definitions of model variables and parameters are shown in Tables A1 - A3.
- **Two - stage stochastic risk - averse programming (TSRP) model**:
- Objective function:
\[
\min f = c^T x + \sum_{h = 1}^{H} p_h q^T y_h + \lambda \left( b + \frac{1}{1 - \alpha} \sum_{h = 1}^{H} p_h v_h \right)
\]
- Constraints:
\[
Ax \leq b
\]
\[
W(w_h) y_h = h(w_h) - T(w_h) x, \quad h = 1, 2, \ldots, H
\]
\[
v_h \geq q^T y_h + c^T x - b, \quad h = 1, 2, \ldots, H
\]
\[
v_h \geq 0, \quad y_h \geq 0, \quad h = 1, 2, \ldots, H
\]
- **Noise - enhanced Arps method**:
- Production decline curve:
\[
Q_t = Q_{i0} \left(1 + b D_i (t - t_0)\right)^{-\frac{1}{b}}
\]
- Cumulative production relationship:
\[
C_p(t) = C_{p0} + \frac{Q_i^b}{D_i (1 - b)} \left[ \left(1 + b D_i (t - t_0)\right)^{\frac{1}{1 - b}} - 1 \right]
\]