Oil & Gas Production Data Analyzer

Minimum data needed for each analysis type:

If minimum data is not available, flag it and show what can still be calculated.

Language Default

Default to Chinese (Simplified) unless the user requests English.

Processing Workflow

Step 1: Parse and Organize Data

Extract columns from the input and create a clean working table:

Identify:

• Time interval (daily / monthly / annual)
• Units (m³, bbl, tons — convert if needed, state conversion used)
• Any missing data points (interpolate linearly if gap ≤ 2 periods, flag if larger)
• Production shutdowns (zero rate periods — do not include in decline fitting)

Step 2: Basic Statistics

Compute:

• Peak production rate and date
• Current production rate (latest data point)
• Cumulative production to date
• Average production rate over period
• Overall decline % from peak to current
• Current water cut and trend (rising / stable / falling)

Step 3: Identify Production Stage

Classify the well or field into a production stage:

Step 4: Decline Curve Analysis (Arps)

Apply only to the declining phase. Do not fit to buildup or plateau periods.

Exponential decline (最常用 for stable reservoir behavior):

q(t) = qi × e^(-Di × t)

Where:

• qi = initial rate at start of decline (m³/d)
• Di = nominal decline rate (1/time)
• t = time since start of decline

Linearize: ln(q) = ln(qi) - Di × t

Estimate Di from slope of ln(q) vs t regression.

Hyperbolic decline (变递减率, most common in practice):

q(t) = qi × (1 + b × Di × t)^(-1/b)

Where b = hyperbolic exponent (0 < b < 1 for most wells)

Estimate by non-linear curve fitting:

• b = 0: reduces to exponential
• b = 1: harmonic decline
• 0.3–0.7: typical for solution gas drive / water drive wells

Report fit quality:

• Show R² for the chosen decline model
• Flag if R² < 0.85 (poor fit, data may have multiple decline segments)

Forecast: Project production to:

1. A user-specified date
2. The economic limit (if provided)
3. 1 year and 5 years from last data point

Show:

预测时间节点 | 预测日产量 (m³/d) | 预测累产 (m³) | 累计递减率 (%)

Step 5: Water Cut Analysis

Method 1: Time-water cut plot (含水率–时间曲线)

• Plot water cut vs time
• Classify trend: linear rise / accelerating / stabilizing

Method 2: Recovery vs Water Cut (甲型水驱特征曲线)

ln(Wo) = a + b × Np

Where Wo = cumulative water production, Np = cumulative oil production.

Linearize on semi-log scale. If linear → water drive is regular.

Forecast:

Np_at_Fw = [(ln(Wo_limit) - a) / b]

Where Wo_limit corresponds to the economic water-oil ratio limit.

Method 3: Water-Oil Ratio vs Cumulative Oil (乙型水驱特征曲线)

WOR = a × e^(b × Np)

Semi-log plot: ln(WOR) vs Np should be linear if regular water drive.

Applicable conditions: relatively homogeneous reservoirs under stable water drive with consistent displacement. For highly heterogeneous or layered reservoirs, prefer 甲型 (Type A) or use both for cross-validation.

State which method was applied and why (data availability, linearity quality).

Step 6: Injection-Production Analysis

If injection data is available:

If Bo and Bw are not provided, use defaults (Bo = 1.2 at initial reservoir pressure for moderate-API crude, Bw = 1.0, Bg calculated from reservoir pressure and temperature if gas data available) and state all assumptions clearly.

Output Format

Production Performance Report

# [Well/Field Name] 生产动态分析报告

> 数据期间：[Start Date] — [End Date]  
> 数据频率：[Daily / Monthly]  
> 分析目的：[Rate trend / Decline / Water cut / Full performance]

---

## 一、生产概况

| 指标 | 数值 |
|------|------|
| 初始产量 (峰值) | m³/d（日期：） |
| 当前产量 | m³/d（日期：） |
| 累计产油量 | m³ |
| 累计产水量 | m³ |
| 当前含水率 | % |
| 生产天数 | d |
| 开采程度（如已知储量）| % |

---

## 二、产量趋势分析

[Narrative describing production trend: buildup, plateau, decline segments, anomalies]

### 生产阶段划分

| 阶段 | 时间段 | 特征 |
|------|--------|------|
| | | |

---

## 三、递减曲线分析

- 递减类型：[指数 / 双曲 / 调和]
- 拟合起始时间：[Date]
- 初始递减率 Di：[value] /月（年递减率：[value]%）
- 双曲指数 b：[value]（适用于双曲递减）
- 拟合 R²：[value]

### 产量预测

| 时间节点 | 预测日产量 (m³/d) | 预测累产量 (m³) |
|---------|----------------|--------------|
| [+1年]  |                |              |
| [+3年]  |                |              |
| [+5年]  |                |              |
| 达到经济极限 |            |              |

---

## 四、含水分析

- 当前含水率：[%]
- 含水上升趋势：[线性上升 / 加速上升 / 趋于稳定]
- 水驱特征曲线类型：[甲型 / 乙型]
- 水驱特征方程：[equation]
- 预测含水 80% 时累计产油量：[m³]
- 预测含水 95% 时累计产油量：[m³]（如有数据支撑）

---

## 五、注采分析（如有注入数据）

| 指标 | 数值 |
|------|------|
| 本期注采比 | |
| 累计注采比 | |
| 地下亏空 / 充填 | m³ |
| 注水见效评价 | |

---

## 六、问题诊断

[List observed anomalies: sudden rate drops, abnormal pressure behavior, unexpected water cut spikes, etc. — with potential causes]

| 现象 | 可能原因 | 建议措施 |
|------|---------|---------|
| | | |

---

## 七、开发建议

[Specific, data-backed recommendations: stimulation, workover, injection adjustment, infill wells, artificial lift optimization]

---

## 八、数据说明与局限性

[Note any data gaps, assumed values, or methodological limitations]

Unit Conversions

If the user provides data in non-standard units, convert and state:

Default oil density for conversion: 0.85 t/m³ (API ~35) — state if different value used.

Engineering Diagnosis Rules

Apply these diagnostic principles automatically when reviewing production data:

Decline Curve Selector

Use this decision guide:

Has the well had consistent decline for ≥ 6 months?
  No → Use trend analysis only, do not fit decline curve
  Yes →
    Is the decline rate relatively constant?
      Yes → Use exponential (b = 0)
      No →
        Is the decline rate decreasing over time?
          Yes → Use hyperbolic (0 < b < 1)
          Is b approaching 1 from fit?
            Hyperbolic b ≈ 1 → Use harmonic

Output Checklist

Before delivering analysis, verify:

• Data table clean and units stated
• Production stage identified
• Basic statistics (peak, current, cumulative) reported
• Decline curve applied only if ≥ 6 months of decline data
• Decline model type stated with R² value
• Forecast table included
• Water cut trend described if water data provided
• Injection-production analysis included if injection data provided
• Problem diagnosis table populated
• Recommendations are data-backed
• Missing data and assumptions stated

Example Prompt Triggers

This skill should activate for prompts such as:

• "这是我们井的月产量数据，帮我分析一下产量递减趋势"
• "帮我做一下这口井的递减曲线分析，预测未来3年产量"
• "含水率一直在涨，帮我分析一下水驱效果"
• "把这个生产数据表整理成标准的生产动态报告"
• "这个注水井的注采比是多少，合理吗"
• "Analyze this production data and fit an Arps decline curve"
• "Forecast production to abandonment for this well"
• "Plot the water cut trend and tell me when we'll hit 90% water cut"
• "Generate a monthly production performance report from this data"
• "用甲型水驱曲线预测这口井的最终采收率"