Add water bills projections

changelog_entry.yaml

@@ -0,0 +1,4 @@
+- bump: minor
+  changes:
+    added:
+    - Water bills projections.

docs/book/_toc.yml

@@ -7,6 +7,7 @@ parts:
   - caption: Economic assumptions
     chapters:
       - file: assumptions/growthfactors
+      - file: assumptions/water-bills
   - caption: Validation
     chapters:
       - file: validation/hbai

docs/book/assumptions/water-bills.md

@@ -0,0 +1,74 @@
+# Water bills projection and uprating
+
+This documentation describes the water bills projection and uprating methodology used in PolicyEngine UK.
+
+## Overview
+
+We implement water bills projections through a combination of historical data analysis and regulatory projection data. The system uses:
+
+1. **Historical data** (2021-2025) from Ofwat average bills
+2. **Regulatory projection data** from Ofwat company-specific increases (2025-2030)
+3. **Economic uprating** using Consumer Price Index (CPIH) growth rates
+
+## Implementation
+
+### Data sources
+
+#### Historical data
+The projection uses historical real water bills data from Ofwat:
+- **2021**: £486 (real terms)
+- **2022**: £470 (real terms)  
+- **2023**: £486 (real terms)
+- **2024**: £492 (real terms)
+- **2025**: £503 (real terms)
+
+**Data source**: [Ofwat average bills data for England and Wales](https://www.ofwat.gov.uk/average-bills-press-statement-2024-25/)
+
+#### Regulatory projections
+We source company-specific water bill increases from the Consumer Council for Water (CCW) based on Ofwat's price review data for 2025-2030. The data includes projected bills for 16 water companies covering both water and sewerage services.
+
+**Data source**: [CCW Water Company Bill Increases 2025-30](https://www.ccw.org.uk/our-work/price-review/how-much-will-my-water-and-sewerage-bills-increase-by-2030/breakdown-of-water-companies-bill-increases-2025-30/)
+
+### Projection methodology
+
+#### Historical period (2021-2025)
+1. We convert real bills to nominal terms using CPIH values
+2. We calculate year-on-year nominal growth rates
+3. We normalize to 2021 baseline (index = 100)
+
+#### Future period (2025-2030)
+1. We calculate average of company-specific projected increases
+2. We apply Consumer Price Index (CPIH) uprating to account for inflation
+3. We compound inflation adjustments year-over-year
+
+The methodology ensures that:
+- We preserve real increases from regulatory decisions
+- We apply additional inflation adjustments using CPIH forecasts
+- Future projections account for both policy changes and economic conditions
+
+### Code structure
+
+#### Core implementation
+- **Location**: `policyengine_uk/utils/water/forecast_water_bills.py`
+- **Function**: `project_water_bills()`
+- **Data**: `policyengine_uk/utils/water/ofwat_increases.csv`
+
+#### Parameters integration
+We store water bills year-on-year growth rates in:
+- **Location**: `policyengine_uk/parameters/gov/economic_assumptions/yoy_growth.yaml`
+- **Parameter**: `ofwat.water_bills`
+
+### Growth rates
+
+The model uses these resulting year-on-year growth rates:
+
+| Year | Growth rate |
+|------|-------------|
+| 2022 | 5.2% |
+| 2023 | 9.2% |
+| 2024 | 4.4% |
+| 2025 | 6.1% |
+| 2026 | 6.1% |
+| 2027 | 5.1% |
+| 2028 | 3.8% |
+| 2029 | 4.3% |

policyengine_uk/data/dataset_schema.py

@@ -86,6 +86,13 @@ def copy(self):
             household=self.household.copy(),
         )
 
+    def validate(self):
+        # Check for NaNs in the tables
+        for df in self.tables:
+            for col in df.columns:
+                if df[col].isna().any():
+                    raise ValueError(f"Column '{col}' contains NaN values.")
+
     @staticmethod
     def from_simulation(
         simulation: "Microsimulation", fiscal_year: int = 2025

policyengine_uk/parameters/gov/economic_assumptions/yoy_growth.yaml

@@ -263,3 +263,21 @@ ons:
       reference:
         - title: ONS Population Projections
           href: https://www.ons.gov.uk/
+ofwat:
+  water_bills:
+    description: Water and sewerage bills year-on-year growth.
+    values:
+      2022-01-01: 0.052
+      2023-01-01: 0.092
+      2024-01-01: 0.044
+      2025-01-01: 0.061
+      2026-01-01: 0.061
+      2027-01-01: 0.051
+      2028-01-01: 0.038
+      2029-01-01: 0.043
+    metadata:
+      unit: /1
+      label: water bills growth
+      reference:
+        - title: Ofwat (and custom projections)
+          href: https://www.ofwat.gov.uk/price-review/

policyengine_uk/utils/water/README.md

@@ -0,0 +1,6 @@
+# Water bills projections
+
+In this folder, we have:
+
+* `forecast_water_bills.py` - A script that projects water bills based on historical data and proposed increases.
+* `ofwat_increases.csv` A CSV with the data from [here](https://www.ccw.org.uk/our-work/price-review/how-much-will-my-water-and-sewerage-bills-increase-by-2030/breakdown-of-water-companies-bill-increases-2025-30/) with proposed pre-inflation increases for each water company.

policyengine_uk/utils/water/forecast_water_bills.py

@@ -0,0 +1,80 @@
+import pandas as pd
+from pathlib import Path
+
+
+def project_water_bills():
+    df_pre_2025 = pd.DataFrame(
+        {
+            "Year": [2021, 2022, 2023, 2024, 2025],
+            "Oftwat avg bills (real)": [486, 470, 486, 492, 503],
+            "CPIH": [113.1, 123.0, 129.9, 134.0, 139.0],
+        }
+    )
+
+    df_pre_2025["Oftwat avg bills (nominal)"] = (
+        df_pre_2025["Oftwat avg bills (real)"] * df_pre_2025["CPIH"] / 100
+    )
+    df_pre_2025["Oftwat avg bills (nominal)"] = (
+        df_pre_2025["Oftwat avg bills (nominal)"]
+        / df_pre_2025["Oftwat avg bills (nominal)"].iloc[0]
+        * 100
+    ).round(1)
+    df_pre_2025["Nominal YoY change"] = (
+        df_pre_2025["Oftwat avg bills (nominal)"].pct_change() * 100
+    ).round(1)
+
+    proposed_increases = pd.read_csv(
+        Path(__file__).parent / "ofwat_increases.csv"
+    )
+    avg_bills_2025_onwards = (
+        proposed_increases[proposed_increases.columns[1:]].mean()[1:].values
+    )
+
+    df_post_2025 = pd.DataFrame(
+        {
+            "Year": [2025, 2026, 2027, 2028, 2029],
+            "CPIH": [139.0, 142.2, 145.2, 148.2, 151.3],
+            "Pre-inflation avg bills (nominal)": avg_bills_2025_onwards,
+        }
+    )
+
+    # Add CPIH to each year's change
+
+    df_post_2025["Avg bills (nominal)"] = df_post_2025[
+        "Pre-inflation avg bills (nominal)"
+    ].values
+    df_post_2025["CPIH change"] = df_post_2025["CPIH"].pct_change() * 100
+
+    for year in range(2026, 2030):
+        row = df_post_2025[df_post_2025["Year"] == year].iloc[0]
+        cpi_change = (
+            row["CPIH"]
+            / df_post_2025[df_post_2025["Year"] == year - 1]["CPIH"].values[0]
+            - 1
+        ) * 100
+        # Increase the nominal bills by the CPIH change
+        addition = df_post_2025.loc[
+            df_post_2025["Year"] == year - 1, "Avg bills (nominal)"
+        ].values[0] * (cpi_change / 100)
+        # Add addition to this and future years
+        df_post_2025.loc[
+            df_post_2025["Year"] >= year, "Avg bills (nominal)"
+        ] += addition
+
+    df_post_2025["Relative change"] = (
+        df_post_2025["Avg bills (nominal)"].pct_change() * 100
+    ).round(1)
+
+    df_post_2025
+
+    combined_water_forecast = pd.DataFrame(
+        {
+            "Year": list(range(2022, 2030)),
+            "Average nominal bills YoY change": df_pre_2025[
+                "Nominal YoY change"
+            ].tolist()[1:]
+            + df_post_2025["Relative change"].tolist()[1:],
+        }
+    )
+
+    print(combined_water_forecast.to_markdown(index=False))

policyengine_uk/variables/input/consumption/property/water_and_sewerage_charges.py

@@ -7,3 +7,4 @@ class water_and_sewerage_charges(Variable):
     label = "water and sewerage charges"
     documentation = "Total amount spent on water and sewerage charges"
     definition_period = YEAR
+    uprating = "gov.economic_assumptions.indices.ofwat.water_bills"