pyconvexity 0.4.8__py3-none-any.whl

pyconvexity/solvers/pypsa/storage.py

@@ -0,0 +1,2096 @@
+"""
+Result storage functionality for PyPSA solver integration.
+
+Handles storing solve results back to the database with proper validation and error handling.
+"""
+
+import uuid
+import pandas as pd
+import numpy as np
+from typing import Dict, Any, Optional, Callable
+
+from pyconvexity.core.types import StaticValue
+from pyconvexity.models import (
+    list_components_by_type,
+    set_static_attribute,
+    set_timeseries_attribute,
+)
+from pyconvexity.validation import get_validation_rule
+
+
+class ResultStorage:
+    """
+    Handles storing PyPSA solve results back to the database.
+
+    This class manages the complex process of extracting results from PyPSA networks
+    and storing them back to the database with proper validation and error handling.
+    """
+
+    def __init__(self, verbose: bool = False):
+        """
+        Initialize ResultStorage.
+
+        Args:
+            verbose: Enable detailed logging output
+        """
+        self.verbose = verbose
+
+    def store_results(
+        self,
+        conn,
+        network: "pypsa.Network",
+        solve_result: Dict[str, Any],
+        scenario_id: Optional[int] = None,
+    ) -> Dict[str, Any]:
+        """
+        Store complete solve results back to database (single network per database).
+
+        Args:
+            conn: Database connection
+            network: Solved PyPSA Network object
+            solve_result: Solve result metadata
+            scenario_id: Optional scenario ID (NULL for base network)
+
+        Returns:
+            Dictionary with storage statistics
+        """
+        run_id = solve_result.get("run_id", str(uuid.uuid4()))
+
+        try:
+            # Store component results
+            component_stats = self._store_component_results(conn, network, scenario_id)
+
+            # Calculate and store clearing prices for all buses
+            clearing_prices_stored = self._calculate_and_store_clearing_prices(
+                conn, network, scenario_id
+            )
+            component_stats["stored_clearing_prices"] = clearing_prices_stored
+
+            # Calculate network statistics first
+            network_stats = self._calculate_network_statistics(
+                conn, network, solve_result
+            )
+
+            # Store solve summary with network statistics
+            self._store_solve_summary(conn, solve_result, scenario_id, network_stats)
+            conn.commit()
+
+            # Store year-based statistics if available
+            year_stats_stored = 0
+            if solve_result.get("year_statistics"):
+                year_stats_stored = self._store_year_based_statistics(
+                    conn, network, solve_result["year_statistics"], scenario_id
+                )
+                conn.commit()
+
+            return {
+                "component_stats": component_stats,
+                "network_stats": network_stats,
+                "year_stats_stored": year_stats_stored,
+                "run_id": run_id,
+                "success": True,
+            }
+
+        except Exception as e:
+            return {
+                "component_stats": {},
+                "network_stats": {},
+                "run_id": run_id,
+                "success": False,
+                "error": str(e),
+            }
+
+    def _store_component_results(
+        self, conn, network: "pypsa.Network", scenario_id: Optional[int]
+    ) -> Dict[str, int]:
+        """Store results for all component types (single network per database)."""
+        results_stats = {
+            "stored_bus_results": 0,
+            "stored_generator_results": 0,
+            "stored_unmet_load_results": 0,
+            "stored_load_results": 0,
+            "stored_line_results": 0,
+            "stored_link_results": 0,
+            "stored_storage_unit_results": 0,
+            "stored_store_results": 0,
+            "skipped_attributes": 0,
+            "errors": 0,
+        }
+
+        try:
+            # Store bus results
+            if hasattr(network, "buses_t") and network.buses_t:
+                results_stats["stored_bus_results"] = (
+                    self._store_component_type_results(
+                        conn, "BUS", network.buses, network.buses_t, scenario_id
+                    )
+                )
+
+            # Store generator results (includes regular generators)
+            if hasattr(network, "generators_t") and network.generators_t:
+                results_stats["stored_generator_results"] = (
+                    self._store_component_type_results(
+                        conn,
+                        "GENERATOR",
+                        network.generators,
+                        network.generators_t,
+                        scenario_id,
+                    )
+                )
+
+                # Store UNMET_LOAD results (these are also stored as generators in PyPSA)
+                results_stats["stored_unmet_load_results"] = (
+                    self._store_component_type_results(
+                        conn,
+                        "UNMET_LOAD",
+                        network.generators,
+                        network.generators_t,
+                        scenario_id,
+                    )
+                )
+
+            # Store load results
+            if hasattr(network, "loads_t") and network.loads_t:
+                results_stats["stored_load_results"] = (
+                    self._store_component_type_results(
+                        conn, "LOAD", network.loads, network.loads_t, scenario_id
+                    )
+                )
+
+            # Store line results
+            if hasattr(network, "lines_t") and network.lines_t:
+                results_stats["stored_line_results"] = (
+                    self._store_component_type_results(
+                        conn, "LINE", network.lines, network.lines_t, scenario_id
+                    )
+                )
+
+            # Store link results
+            if hasattr(network, "links_t") and network.links_t:
+                results_stats["stored_link_results"] = (
+                    self._store_component_type_results(
+                        conn, "LINK", network.links, network.links_t, scenario_id
+                    )
+                )
+
+            # Store storage unit results
+            if hasattr(network, "storage_units_t") and network.storage_units_t:
+                results_stats["stored_storage_unit_results"] = (
+                    self._store_component_type_results(
+                        conn,
+                        "STORAGE_UNIT",
+                        network.storage_units,
+                        network.storage_units_t,
+                        scenario_id,
+                    )
+                )
+
+            # Store store results
+            if hasattr(network, "stores_t") and network.stores_t:
+                results_stats["stored_store_results"] = (
+                    self._store_component_type_results(
+                        conn, "STORE", network.stores, network.stores_t, scenario_id
+                    )
+                )
+
+            return results_stats
+
+        except Exception as e:
+            results_stats["errors"] += 1
+            return results_stats
+
+    def _calculate_and_store_clearing_prices(
+        self,
+        conn,
+        network: "pypsa.Network",
+        scenario_id: Optional[int],
+    ) -> int:
+        """
+        Calculate and store clearing prices for all buses.
+        
+        The clearing price at each bus is the pay-as-clear price: the marginal
+        cost of the cheapest source (generator, storage, or import) with spare
+        capacity. This differs from the marginal_price (LP shadow price).
+        
+        Args:
+            conn: Database connection
+            network: Solved PyPSA Network object
+            scenario_id: Scenario ID for result storage
+            
+        Returns:
+            Number of buses with clearing prices stored
+        """
+        logger.info(f"=== CALCULATING AND STORING CLEARING PRICES (scenario_id={scenario_id}) ===")
+        
+        try:
+            from .clearing_price import ClearingPriceCalculator
+            
+            calculator = ClearingPriceCalculator(verbose=True)
+            clearing_prices = calculator.calculate_all_buses(conn, network, scenario_id)
+            
+            if not clearing_prices:
+                logger.warning("No clearing prices calculated - clearing_prices dict is empty")
+                return 0
+            
+            # Log what we got from the calculator
+            logger.info(f"Clearing prices calculated for {len(clearing_prices)} buses")
+            for bus_name, prices in clearing_prices.items():
+                n_zeros = np.sum(prices == 0)
+                n_inf = np.sum(np.isinf(prices))
+                valid = prices[(prices > 0) & np.isfinite(prices)]
+                if len(valid) > 0:
+                    logger.info(f"  {bus_name}: {len(prices)} periods, mean=£{np.mean(valid):.2f}, "
+                               f"zeros={n_zeros}, inf={n_inf}, range=[£{np.min(valid):.2f}, £{np.max(valid):.2f}]")
+                else:
+                    logger.warning(f"  {bus_name}: {len(prices)} periods, ALL ZERO OR INF (zeros={n_zeros}, inf={n_inf})")
+            
+            # Get bus component IDs
+            buses = list_components_by_type(conn, "BUS")
+            bus_name_to_id = {bus.name: bus.id for bus in buses}
+            logger.info(f"Found {len(buses)} buses in database: {list(bus_name_to_id.keys())}")
+            
+            stored_count = 0
+            for bus_name, prices in clearing_prices.items():
+                if bus_name not in bus_name_to_id:
+                    logger.warning(f"  {bus_name}: not found in database - skipping")
+                    continue
+                
+                bus_id = bus_name_to_id[bus_name]
+                values = [float(p) if np.isfinite(p) else 0.0 for p in prices]
+                
+                # Log what we're about to store
+                n_zeros = sum(1 for v in values if v == 0)
+                if n_zeros > 0:
+                    logger.warning(f"  {bus_name}: storing {len(values)} values, {n_zeros} zeros")
+                
+                try:
+                    set_timeseries_attribute(
+                        conn, bus_id, "clearing_price", values, scenario_id
+                    )
+                    stored_count += 1
+                    logger.info(f"  ✅ {bus_name} (id={bus_id}): stored {len(values)} clearing prices")
+                except Exception as e:
+                    logger.error(f"  ❌ {bus_name} (id={bus_id}): failed to store clearing_price: {e}")
+                    import traceback
+                    traceback.print_exc()
+                    continue
+            
+            logger.info(f"=== CLEARING PRICES: Stored {stored_count}/{len(clearing_prices)} buses ===")
+            return stored_count
+            
+        except ImportError as e:
+            logger.error(f"ClearingPriceCalculator not available - skipping clearing price calculation: {e}")
+            return 0
+        except Exception as e:
+            logger.error(f"Failed to calculate/store clearing prices: {e}")
+            import traceback
+            traceback.print_exc()
+            return 0
+
+    def _store_component_type_results(
+        self,
+        conn,
+        component_type: str,
+        static_df: pd.DataFrame,
+        timeseries_dict: Dict[str, pd.DataFrame],
+        scenario_id: Optional[int],
+    ) -> int:
+        """Store results for a specific component type - only store OUTPUT attributes (single network per database)."""
+        stored_count = 0
+
+        try:
+            # Get component name to ID mapping
+            components = list_components_by_type(conn, component_type)
+            name_to_id = {comp.name: comp.id for comp in components}
+
+            # Store timeseries results - ONLY OUTPUT ATTRIBUTES (is_input=FALSE)
+            for attr_name, timeseries_df in timeseries_dict.items():
+                if timeseries_df.empty:
+                    continue
+
+                # Check if this attribute is an output attribute (not an input)
+                try:
+                    rule = get_validation_rule(conn, component_type, attr_name)
+                    if rule.is_input:
+                        # Skip input attributes to preserve original input data
+                        continue
+                except Exception:
+                    # If no validation rule found, skip to be safe
+                    continue
+
+                for component_name in timeseries_df.columns:
+                    if component_name not in name_to_id:
+                        continue
+
+                    component_id = name_to_id[component_name]
+                    component_series = timeseries_df[component_name]
+
+                    # Skip if all values are NaN
+                    if component_series.isna().all():
+                        continue
+
+                    # Convert to efficient values array
+                    values = []
+                    for value in component_series.values:
+                        if pd.isna(value):
+                            values.append(0.0)  # Fill NaN with 0.0
+                        else:
+                            values.append(float(value))
+
+                    if not values:
+                        continue
+
+                    # Store using efficient format
+                    try:
+                        set_timeseries_attribute(
+                            conn, component_id, attr_name, values, scenario_id
+                        )
+                        stored_count += 1
+                    except Exception as e:
+                        # Handle validation errors gracefully
+                        if (
+                            "No validation rule found" in str(e)
+                            or "does not allow" in str(e)
+                            or "ValidationError" in str(type(e).__name__)
+                        ):
+                            continue
+                        else:
+                            continue
+
+            # Store static optimization results - ONLY OUTPUT ATTRIBUTES (is_input=FALSE)
+            if not static_df.empty:
+                for attr_name in static_df.columns:
+                    # Check if this attribute is an output attribute (not an input)
+                    try:
+                        rule = get_validation_rule(conn, component_type, attr_name)
+                        if rule.is_input:
+                            # Skip input attributes to preserve original input data
+                            continue
+                    except Exception:
+                        # If no validation rule found, skip to be safe
+                        continue
+
+                    for component_name, value in static_df[attr_name].items():
+                        if component_name not in name_to_id or pd.isna(value):
+                            continue
+
+                        component_id = name_to_id[component_name]
+
+                        try:
+                            # Convert value to StaticValue
+                            if isinstance(value, (int, np.integer)):
+                                static_value = StaticValue(int(value))
+                            elif isinstance(value, (float, np.floating)):
+                                if np.isfinite(value):
+                                    static_value = StaticValue(float(value))
+                                else:
+                                    continue  # Skip infinite/NaN values
+                            elif isinstance(value, bool):
+                                static_value = StaticValue(bool(value))
+                            else:
+                                static_value = StaticValue(str(value))
+
+                            # Store using atomic utility
+                            set_static_attribute(
+                                conn, component_id, attr_name, static_value, scenario_id
+                            )
+                            stored_count += 1
+
+                        except Exception as e:
+                            # Handle validation errors gracefully
+                            if (
+                                "No validation rule found" in str(e)
+                                or "does not allow" in str(e)
+                                or "ValidationError" in str(type(e).__name__)
+                            ):
+                                continue
+                            else:
+                                continue
+
+            return stored_count
+
+        except Exception as e:
+            return stored_count
+
+    def _store_solve_summary(
+        self,
+        conn,
+        solve_result: Dict[str, Any],
+        scenario_id: Optional[int],
+        network_stats: Optional[Dict[str, Any]] = None,
+    ):
+        """Store solve summary to network_solve_results table (single network per database)."""
+        try:
+            # Prepare solve summary data
+            solver_name = solve_result.get("solver_name", "unknown")
+            solve_status = solve_result.get("status", "unknown")
+            objective_value = solve_result.get("objective_value")
+            solve_time = solve_result.get("solve_time", 0.0)
+
+            # Create enhanced solve result with network statistics for serialization
+            enhanced_solve_result = {
+                **solve_result,
+                "network_statistics": network_stats or {},
+            }
+
+            # Delete existing result for this scenario first (handles NULL scenario_id correctly)
+            if scenario_id is None:
+                conn.execute(
+                    "DELETE FROM network_solve_results WHERE scenario_id IS NULL"
+                )
+            else:
+                conn.execute(
+                    "DELETE FROM network_solve_results WHERE scenario_id = ?",
+                    (scenario_id,),
+                )
+
+            results_json = self._serialize_results_json(enhanced_solve_result)
+            metadata_json = self._serialize_metadata_json(enhanced_solve_result)
+
+            # Insert new solve results summary
+            conn.execute(
+                """
+                INSERT INTO network_solve_results (
+                    scenario_id, solver_name, solve_type, solve_status,
+                    objective_value, solve_time_seconds, results_json, metadata_json
+                ) VALUES (?, ?, ?, ?, ?, ?, ?, ?)
+            """,
+                (
+                    scenario_id,
+                    solver_name,
+                    "pypsa_optimization",
+                    solve_status,
+                    objective_value,
+                    solve_time,
+                    results_json,
+                    metadata_json,
+                ),
+            )
+
+        except Exception as e:
+            raise  # Re-raise to trigger rollback
+
+    def _calculate_network_statistics(
+        self, conn, network: "pypsa.Network", solve_result: Dict[str, Any]
+    ) -> Dict[str, Any]:
+        """Calculate network statistics - focusing only on capacity for now (single network per database)."""
+        try:
+            # Calculate carrier-specific statistics
+            carrier_stats = self._calculate_carrier_statistics(conn, network)
+
+            # Calculate basic network statistics
+            total_cost = solve_result.get("objective_value", 0.0)
+            total_generation_mwh = sum(
+                carrier_stats.get("dispatch_by_carrier", {}).values()
+            )
+            total_emissions_tonnes = sum(
+                carrier_stats.get("emissions_by_carrier", {}).values()
+            )
+            total_capital_cost = sum(
+                carrier_stats.get("capital_cost_by_carrier", {}).values()
+            )
+            total_operational_cost = sum(
+                carrier_stats.get("operational_cost_by_carrier", {}).values()
+            )
+            total_system_cost = sum(
+                carrier_stats.get("total_system_cost_by_carrier", {}).values()
+            )
+
+            # Calculate unmet load statistics
+            unmet_load_mwh = carrier_stats.get("dispatch_by_carrier", {}).get(
+                "Unmet Load", 0.0
+            )
+            total_demand_mwh = self._calculate_total_demand(network)
+            unmet_load_percentage = (
+                (unmet_load_mwh / (total_demand_mwh + 1e-6)) * 100
+                if total_demand_mwh > 0
+                else 0.0
+            )
+
+            # Create nested structure expected by frontend
+            network_statistics = {
+                "core_summary": {
+                    "total_generation_mwh": total_generation_mwh,
+                    "total_demand_mwh": total_demand_mwh,
+                    "total_cost": total_cost,
+                    "load_factor": (
+                        (total_demand_mwh / (total_generation_mwh + 1e-6))
+                        if total_generation_mwh > 0
+                        else 0.0
+                    ),
+                    "unserved_energy_mwh": unmet_load_mwh,
+                },
+                "custom_statistics": {
+                    # Include carrier-specific statistics (capacity, dispatch, emissions, costs)
+                    **carrier_stats,
+                    "total_capital_cost": total_capital_cost,
+                    "total_operational_cost": total_operational_cost,
+                    "total_currency_cost": total_system_cost,  # Use calculated system cost instead of PyPSA objective
+                    "total_emissions_tons_co2": total_emissions_tonnes,
+                    "average_price_per_mwh": (
+                        (total_system_cost / (total_generation_mwh + 1e-6))
+                        if total_generation_mwh > 0
+                        else 0.0
+                    ),
+                    "unmet_load_percentage": unmet_load_percentage,
+                    "max_unmet_load_hour_mw": 0.0,  # TODO: Calculate max hourly unmet load later
+                },
+                "runtime_info": {
+                    "component_count": (
+                        (
+                            len(network.buses)
+                            + len(network.generators)
+                            + len(network.loads)
+                            + len(network.lines)
+                            + len(network.links)
+                        )
+                        if hasattr(network, "buses")
+                        else 0
+                    ),
+                    "bus_count": len(network.buses) if hasattr(network, "buses") else 0,
+                    "generator_count": (
+                        len(network.generators) if hasattr(network, "generators") else 0
+                    ),
+                    "load_count": (
+                        len(network.loads) if hasattr(network, "loads") else 0
+                    ),
+                    "snapshot_count": (
+                        len(network.snapshots) if hasattr(network, "snapshots") else 0
+                    ),
+                },
+            }
+
+            return network_statistics
+
+        except Exception as e:
+            # Return empty structure matching expected format
+            return {
+                "core_summary": {
+                    "total_generation_mwh": 0.0,
+                    "total_demand_mwh": 0.0,
+                    "total_cost": solve_result.get("objective_value", 0.0),
+                    "load_factor": 0.0,
+                    "unserved_energy_mwh": 0.0,
+                },
+                "custom_statistics": {
+                    "dispatch_by_carrier": {},
+                    "power_capacity_by_carrier": {},
+                    "energy_capacity_by_carrier": {},
+                    "emissions_by_carrier": {},
+                    "capital_cost_by_carrier": {},
+                    "operational_cost_by_carrier": {},
+                    "total_system_cost_by_carrier": {},
+                    "total_capital_cost": 0.0,
+                    "total_operational_cost": 0.0,
+                    "total_currency_cost": 0.0,
+                    "total_emissions_tons_co2": 0.0,
+                    "average_price_per_mwh": 0.0,
+                    "unmet_load_percentage": 0.0,
+                    "max_unmet_load_hour_mw": 0.0,
+                },
+                "runtime_info": {
+                    "component_count": 0,
+                    "bus_count": 0,
+                    "generator_count": 0,
+                    "load_count": 0,
+                    "snapshot_count": 0,
+                },
+                "error": str(e),
+            }
+
+    def _calculate_carrier_statistics(
+        self, conn, network: "pypsa.Network"
+    ) -> Dict[str, Any]:
+        """
+        Calculate carrier-specific statistics directly from the network (single network per database).
+        This is the primary calculation - per-year stats will be calculated separately.
+        """
+        try:
+            # Calculate all-year statistics directly from the network
+            # Extract years from network snapshots
+            if hasattr(network.snapshots, "levels"):
+                # Multi-period optimization - get years from period level
+                period_values = network.snapshots.get_level_values(0)
+                years = sorted(period_values.unique())
+            elif hasattr(network.snapshots, "year"):
+                years = sorted(network.snapshots.year.unique())
+            elif hasattr(network, "_available_years"):
+                years = network._available_years
+            else:
+                years = [2020]  # Fallback
+
+            # Calculate per-year statistics first
+            all_year_stats = {
+                "dispatch_by_carrier": {},
+                "power_capacity_by_carrier": {},
+                "energy_capacity_by_carrier": {},
+                "emissions_by_carrier": {},
+                "capital_cost_by_carrier": {},
+                "operational_cost_by_carrier": {},
+                "total_system_cost_by_carrier": {},
+            }
+
+            # Initialize all carriers with zero values
+            cursor = conn.execute(
+                """
+                SELECT DISTINCT name FROM carriers
+            """
+            )
+            all_carriers = [row[0] for row in cursor.fetchall()]
+
+            # Initialize all carriers with zero values (including special "Unmet Load" carrier)
+            all_carriers_with_unmet = all_carriers + ["Unmet Load"]
+            for carrier in all_carriers_with_unmet:
+                all_year_stats["dispatch_by_carrier"][carrier] = 0.0
+                all_year_stats["power_capacity_by_carrier"][carrier] = 0.0
+                all_year_stats["energy_capacity_by_carrier"][carrier] = 0.0
+                all_year_stats["emissions_by_carrier"][carrier] = 0.0
+                all_year_stats["capital_cost_by_carrier"][carrier] = 0.0
+                all_year_stats["operational_cost_by_carrier"][carrier] = 0.0
+                all_year_stats["total_system_cost_by_carrier"][carrier] = 0.0
+
+            # Calculate statistics for each year and sum them up
+            for year in years:
+                year_stats = self._calculate_year_carrier_statistics(
+                    conn, network, year
+                )
+
+                # Sum up all the statistics (including "Unmet Load")
+                for carrier in all_carriers_with_unmet:
+                    # Sum dispatch, emissions, and costs across years
+                    all_year_stats["dispatch_by_carrier"][carrier] += year_stats[
+                        "dispatch_by_carrier"
+                    ].get(carrier, 0.0)
+                    all_year_stats["emissions_by_carrier"][carrier] += year_stats[
+                        "emissions_by_carrier"
+                    ].get(carrier, 0.0)
+                    all_year_stats["capital_cost_by_carrier"][carrier] += year_stats[
+                        "capital_cost_by_carrier"
+                    ].get(carrier, 0.0)
+                    all_year_stats["operational_cost_by_carrier"][
+                        carrier
+                    ] += year_stats["operational_cost_by_carrier"].get(carrier, 0.0)
+                    all_year_stats["total_system_cost_by_carrier"][
+                        carrier
+                    ] += year_stats["total_system_cost_by_carrier"].get(carrier, 0.0)
+
+                    # For capacity: use the last year (final capacity state)
+                    if year == years[-1]:
+                        all_year_stats["power_capacity_by_carrier"][carrier] = (
+                            year_stats["power_capacity_by_carrier"].get(carrier, 0.0)
+                        )
+                        all_year_stats["energy_capacity_by_carrier"][carrier] = (
+                            year_stats["energy_capacity_by_carrier"].get(carrier, 0.0)
+                        )
+
+            return all_year_stats
+
+        except Exception as e:
+            return {
+                "dispatch_by_carrier": {},
+                "power_capacity_by_carrier": {},
+                "energy_capacity_by_carrier": {},
+                "emissions_by_carrier": {},
+                "capital_cost_by_carrier": {},
+                "operational_cost_by_carrier": {},
+                "total_system_cost_by_carrier": {},
+            }
+
+    def _store_year_based_statistics(
+        self,
+        conn,
+        network: "pypsa.Network",
+        year_statistics: Dict[int, Dict[str, Any]],
+        scenario_id: Optional[int],
+    ) -> int:
+        """Store year-based statistics to database (single network per database)"""
+        try:
+            import json
+
+            stored_count = 0
+
+            # Check if network_solve_results_by_year table exists, create if not
+            conn.execute(
+                """
+                CREATE TABLE IF NOT EXISTS network_solve_results_by_year (
+                    id INTEGER PRIMARY KEY AUTOINCREMENT,
+                    scenario_id INTEGER,
+                    year INTEGER NOT NULL,
+                    results_json TEXT,
+                    metadata_json TEXT,
+                    created_at DATETIME DEFAULT CURRENT_TIMESTAMP,
+                    FOREIGN KEY (scenario_id) REFERENCES scenarios(id),
+                    UNIQUE(scenario_id, year)
+                )
+            """
+            )
+
+            for year, stats in year_statistics.items():
+                try:
+                    # Calculate proper year-specific carrier statistics
+                    year_carrier_stats = self._calculate_year_carrier_statistics(
+                        conn, network, year
+                    )
+
+                    # Merge year-specific carrier stats into the statistics
+                    if "custom_statistics" in stats:
+                        stats["custom_statistics"].update(year_carrier_stats)
+                    else:
+                        stats["custom_statistics"] = year_carrier_stats
+
+                    # Wrap the year statistics in the same structure as overall results for consistency
+                    year_result_wrapper = {
+                        "success": True,
+                        "year": year,
+                        "network_statistics": stats,
+                    }
+
+                    metadata = {"year": year, "scenario_id": scenario_id}
+
+                    # Delete existing result for this scenario+year first (handles NULL scenario_id correctly)
+                    if scenario_id is None:
+                        conn.execute(
+                            """
+                            DELETE FROM network_solve_results_by_year 
+                            WHERE scenario_id IS NULL AND year = ?
+                        """,
+                            (year,),
+                        )
+                    else:
+                        conn.execute(
+                            """
+                            DELETE FROM network_solve_results_by_year 
+                            WHERE scenario_id = ? AND year = ?
+                        """,
+                            (scenario_id, year),
+                        )
+
+                    # Insert new year-based results
+                    conn.execute(
+                        """
+                        INSERT INTO network_solve_results_by_year 
+                        (scenario_id, year, results_json, metadata_json)
+                        VALUES (?, ?, ?, ?)
+                    """,
+                        (
+                            scenario_id,
+                            year,
+                            json.dumps(
+                                year_result_wrapper, default=self._json_serializer
+                            ),
+                            json.dumps(metadata, default=self._json_serializer),
+                        ),
+                    )
+
+                    stored_count += 1
+
+                except Exception as e:
+                    continue
+
+            return stored_count
+
+        except Exception as e:
+            return 0
+
+    def _calculate_year_carrier_statistics(
+        self, conn, network: "pypsa.Network", year: int
+    ) -> Dict[str, Any]:
+        """
+        Calculate carrier-specific statistics for a specific year.
+        For now, only calculate capacity statistics.
+        """
+        try:
+            # Initialize carrier statistics
+            carrier_stats = {
+                "dispatch_by_carrier": {},
+                "power_capacity_by_carrier": {},  # MW - Generators + Storage Units (power) + Lines + Links
+                "energy_capacity_by_carrier": {},  # MWh - Stores + Storage Units (energy)
+                "emissions_by_carrier": {},
+                "capital_cost_by_carrier": {},
+                "operational_cost_by_carrier": {},
+                "total_system_cost_by_carrier": {},
+            }
+
+            # Get all carriers from database
+            cursor = conn.execute(
+                """
+                SELECT DISTINCT name FROM carriers
+            """
+            )
+            all_carriers = [row[0] for row in cursor.fetchall()]
+
+            # Initialize all carriers with zero values (including special "Unmet Load" carrier)
+            all_carriers_with_unmet = all_carriers + ["Unmet Load"]
+            for carrier in all_carriers_with_unmet:
+                carrier_stats["dispatch_by_carrier"][carrier] = 0.0
+                carrier_stats["power_capacity_by_carrier"][carrier] = 0.0
+                carrier_stats["energy_capacity_by_carrier"][carrier] = 0.0
+                carrier_stats["emissions_by_carrier"][carrier] = 0.0
+                carrier_stats["capital_cost_by_carrier"][carrier] = 0.0
+                carrier_stats["operational_cost_by_carrier"][carrier] = 0.0
+                carrier_stats["total_system_cost_by_carrier"][carrier] = 0.0
+
+            # Calculate dispatch (generation) by carrier for this specific year
+
+            # 1. GENERATORS - Generation dispatch (including UNMET_LOAD)
+            if hasattr(network, "generators_t") and hasattr(network.generators_t, "p"):
+                # Get generator-carrier mapping (include both GENERATOR and UNMET_LOAD)
+                cursor = conn.execute(
+                    """
+                SELECT c.name as component_name, 
+                       CASE 
+                           WHEN c.component_type = 'UNMET_LOAD' THEN 'Unmet Load'
+                           ELSE carr.name 
+                       END as carrier_name
+                FROM components c
+                JOIN carriers carr ON c.carrier_id = carr.id
+                WHERE c.component_type IN ('GENERATOR', 'UNMET_LOAD')
+            """
+                )
+                generator_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                # Filter generation data for this specific year
+                year_generation = self._filter_timeseries_by_year(
+                    network.generators_t.p, network.snapshots, year
+                )
+                if year_generation is not None and not year_generation.empty:
+                    for gen_name in year_generation.columns:
+                        if gen_name in generator_carriers:
+                            carrier_name = generator_carriers[gen_name]
+
+                            # Calculate generation for this year (ALWAYS apply snapshot weightings to convert MW to MWh)
+                            year_weightings = self._get_year_weightings(network, year)
+                            if year_weightings is not None:
+                                generation_mwh = float(
+                                    (
+                                        year_generation[gen_name].values
+                                        * year_weightings
+                                    ).sum()
+                                )
+                            else:
+                                # Fallback: simple sum (will be incorrect for non-1H models)
+                                generation_mwh = float(year_generation[gen_name].sum())
+
+                            if carrier_name in carrier_stats["dispatch_by_carrier"]:
+                                carrier_stats["dispatch_by_carrier"][
+                                    carrier_name
+                                ] += generation_mwh
+
+            # 2. STORAGE_UNITS - Discharge only (positive values)
+            if hasattr(network, "storage_units_t") and hasattr(
+                network.storage_units_t, "p"
+            ):
+                # Get storage unit-carrier mapping
+                cursor = conn.execute(
+                    """
+                    SELECT c.name as component_name, carr.name as carrier_name
+                    FROM components c
+                    JOIN carriers carr ON c.carrier_id = carr.id
+                    WHERE c.component_type = 'STORAGE_UNIT'
+                """
+                )
+                storage_unit_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                # Filter storage unit data for this specific year
+                year_storage = self._filter_timeseries_by_year(
+                    network.storage_units_t.p, network.snapshots, year
+                )
+                if year_storage is not None and not year_storage.empty:
+                    for su_name in year_storage.columns:
+                        if su_name in storage_unit_carriers:
+                            carrier_name = storage_unit_carriers[su_name]
+
+                            # Calculate discharge for this year (positive values only, ALWAYS apply snapshot weightings)
+                            year_weightings = self._get_year_weightings(network, year)
+                            if year_weightings is not None:
+                                discharge_mwh = float(
+                                    (
+                                        year_storage[su_name].clip(lower=0).values
+                                        * year_weightings
+                                    ).sum()
+                                )
+                            else:
+                                # Fallback: simple sum (will be incorrect for non-1H models)
+                                discharge_mwh = float(
+                                    year_storage[su_name].clip(lower=0).sum()
+                                )
+
+                            if carrier_name in carrier_stats["dispatch_by_carrier"]:
+                                carrier_stats["dispatch_by_carrier"][
+                                    carrier_name
+                                ] += discharge_mwh
+
+            # 3. STORES - Discharge only (positive values)
+            if hasattr(network, "stores_t") and hasattr(network.stores_t, "p"):
+                # Get store-carrier mapping
+                cursor = conn.execute(
+                    """
+                    SELECT c.name as component_name, carr.name as carrier_name
+                    FROM components c
+                    JOIN carriers carr ON c.carrier_id = carr.id
+                    WHERE c.component_type = 'STORE'
+                """
+                )
+                store_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                # Filter store data for this specific year
+                year_stores = self._filter_timeseries_by_year(
+                    network.stores_t.p, network.snapshots, year
+                )
+                if year_stores is not None and not year_stores.empty:
+                    for store_name in year_stores.columns:
+                        if store_name in store_carriers:
+                            carrier_name = store_carriers[store_name]
+
+                            # Calculate discharge for this year (positive values only, ALWAYS apply snapshot weightings)
+                            year_weightings = self._get_year_weightings(network, year)
+                            if year_weightings is not None:
+                                discharge_mwh = float(
+                                    (
+                                        year_stores[store_name].clip(lower=0).values
+                                        * year_weightings
+                                    ).sum()
+                                )
+                            else:
+                                # Fallback: simple sum (will be incorrect for non-1H models)
+                                discharge_mwh = float(
+                                    year_stores[store_name].clip(lower=0).sum()
+                                )
+
+                            if carrier_name in carrier_stats["dispatch_by_carrier"]:
+                                carrier_stats["dispatch_by_carrier"][
+                                    carrier_name
+                                ] += discharge_mwh
+
+            # Calculate emissions by carrier for this specific year
+            # Get emission factors for all carriers
+            cursor = conn.execute(
+                """
+                SELECT name, co2_emissions FROM carriers
+            """
+            )
+            emission_factors = {row[0]: row[1] for row in cursor.fetchall()}
+
+            # Calculate emissions: dispatch (MWh) × emission factor (tonnes CO2/MWh) = tonnes CO2
+            for carrier_name, dispatch_mwh in carrier_stats[
+                "dispatch_by_carrier"
+            ].items():
+                # Handle None values safely
+                if dispatch_mwh is None:
+                    dispatch_mwh = 0.0
+
+                emission_factor = emission_factors.get(
+                    carrier_name, 0.0
+                )  # Default to 0 if no emission factor
+                if emission_factor is None:
+                    emission_factor = 0.0
+
+                emissions_tonnes = dispatch_mwh * emission_factor
+
+                if carrier_name in carrier_stats["emissions_by_carrier"]:
+                    carrier_stats["emissions_by_carrier"][
+                        carrier_name
+                    ] += emissions_tonnes
+
+            # Calculate capital costs by carrier for this specific year
+            # Capital costs are annualized and counted every year the component is active
+
+            # Helper function to check if component is active in this year
+            def is_component_active(build_year, lifetime, current_year):
+                """Check if component is active in the current year based on build_year and lifetime"""
+                if pd.isna(build_year):
+                    return True  # No build year constraint
+
+                build_year = int(build_year)
+                if build_year > current_year:
+                    return False  # Not built yet
+
+                if pd.isna(lifetime) or lifetime == float("inf"):
+                    return True  # Infinite lifetime
+
+                lifetime = int(lifetime)
+                end_year = build_year + lifetime - 1
+                return current_year <= end_year
+
+            # 1. GENERATORS - Capital costs (including UNMET_LOAD)
+            if hasattr(network, "generators") and not network.generators.empty:
+                # Get generator info: carrier, capital_cost, build_year, lifetime (include UNMET_LOAD)
+                cursor = conn.execute(
+                    """
+                    SELECT c.name as component_name, 
+                           CASE 
+                               WHEN c.component_type = 'UNMET_LOAD' THEN 'Unmet Load'
+                               ELSE carr.name 
+                           END as carrier_name
+                    FROM components c
+                    JOIN carriers carr ON c.carrier_id = carr.id
+                    WHERE c.component_type IN ('GENERATOR', 'UNMET_LOAD')
+                """
+                )
+                generator_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                for gen_name in network.generators.index:
+                    if gen_name in generator_carriers:
+                        carrier_name = generator_carriers[gen_name]
+
+                        # Get build year and lifetime
+                        build_year = (
+                            network.generators.loc[gen_name, "build_year"]
+                            if "build_year" in network.generators.columns
+                            else None
+                        )
+                        lifetime = (
+                            network.generators.loc[gen_name, "lifetime"]
+                            if "lifetime" in network.generators.columns
+                            else None
+                        )
+
+                        # Check if component is active in this year
+                        if is_component_active(build_year, lifetime, year):
+                            # Get capacity and capital cost
+                            if "p_nom_opt" in network.generators.columns:
+                                capacity_mw = float(
+                                    network.generators.loc[gen_name, "p_nom_opt"]
+                                )
+                            else:
+                                capacity_mw = (
+                                    float(network.generators.loc[gen_name, "p_nom"])
+                                    if "p_nom" in network.generators.columns
+                                    else 0.0
+                                )
+
+                            capital_cost_per_mw = (
+                                float(network.generators.loc[gen_name, "capital_cost"])
+                                if "capital_cost" in network.generators.columns
+                                else 0.0
+                            )
+
+                            # Calculate annualized capital cost for this year
+                            annual_capital_cost = capacity_mw * capital_cost_per_mw
+
+                            if carrier_name in carrier_stats["capital_cost_by_carrier"]:
+                                carrier_stats["capital_cost_by_carrier"][
+                                    carrier_name
+                                ] += annual_capital_cost
+
+            # 2. STORAGE_UNITS - Capital costs
+            if hasattr(network, "storage_units") and not network.storage_units.empty:
+                cursor = conn.execute(
+                    """
+                    SELECT c.name as component_name, carr.name as carrier_name
+                    FROM components c
+                    JOIN carriers carr ON c.carrier_id = carr.id
+                    WHERE c.component_type = 'STORAGE_UNIT'
+                """
+                )
+                storage_unit_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                for su_name in network.storage_units.index:
+                    if su_name in storage_unit_carriers:
+                        carrier_name = storage_unit_carriers[su_name]
+
+                        # Get build year and lifetime
+                        build_year = (
+                            network.storage_units.loc[su_name, "build_year"]
+                            if "build_year" in network.storage_units.columns
+                            else None
+                        )
+                        lifetime = (
+                            network.storage_units.loc[su_name, "lifetime"]
+                            if "lifetime" in network.storage_units.columns
+                            else None
+                        )
+
+                        # Check if component is active in this year
+                        if is_component_active(build_year, lifetime, year):
+                            # Get power capacity and capital cost (per MW)
+                            if "p_nom_opt" in network.storage_units.columns:
+                                capacity_mw = float(
+                                    network.storage_units.loc[su_name, "p_nom_opt"]
+                                )
+                            else:
+                                capacity_mw = (
+                                    float(network.storage_units.loc[su_name, "p_nom"])
+                                    if "p_nom" in network.storage_units.columns
+                                    else 0.0
+                                )
+
+                            capital_cost_per_mw = (
+                                float(
+                                    network.storage_units.loc[su_name, "capital_cost"]
+                                )
+                                if "capital_cost" in network.storage_units.columns
+                                else 0.0
+                            )
+
+                            # Calculate annualized capital cost for this year
+                            annual_capital_cost = capacity_mw * capital_cost_per_mw
+
+                            if carrier_name in carrier_stats["capital_cost_by_carrier"]:
+                                carrier_stats["capital_cost_by_carrier"][
+                                    carrier_name
+                                ] += annual_capital_cost
+
+            # 3. STORES - Capital costs (per MWh)
+            if hasattr(network, "stores") and not network.stores.empty:
+                cursor = conn.execute(
+                    """
+                    SELECT c.name as component_name, carr.name as carrier_name
+                    FROM components c
+                    JOIN carriers carr ON c.carrier_id = carr.id
+                    WHERE c.component_type = 'STORE'
+                """
+                )
+                store_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                for store_name in network.stores.index:
+                    if store_name in store_carriers:
+                        carrier_name = store_carriers[store_name]
+
+                        # Get build year and lifetime
+                        build_year = (
+                            network.stores.loc[store_name, "build_year"]
+                            if "build_year" in network.stores.columns
+                            else None
+                        )
+                        lifetime = (
+                            network.stores.loc[store_name, "lifetime"]
+                            if "lifetime" in network.stores.columns
+                            else None
+                        )
+
+                        # Check if component is active in this year
+                        if is_component_active(build_year, lifetime, year):
+                            # Get energy capacity and capital cost (per MWh)
+                            if "e_nom_opt" in network.stores.columns:
+                                capacity_mwh = float(
+                                    network.stores.loc[store_name, "e_nom_opt"]
+                                )
+                            else:
+                                capacity_mwh = (
+                                    float(network.stores.loc[store_name, "e_nom"])
+                                    if "e_nom" in network.stores.columns
+                                    else 0.0
+                                )
+
+                            capital_cost_per_mwh = (
+                                float(network.stores.loc[store_name, "capital_cost"])
+                                if "capital_cost" in network.stores.columns
+                                else 0.0
+                            )
+
+                            # Calculate annualized capital cost for this year
+                            annual_capital_cost = capacity_mwh * capital_cost_per_mwh
+
+                            if carrier_name in carrier_stats["capital_cost_by_carrier"]:
+                                carrier_stats["capital_cost_by_carrier"][
+                                    carrier_name
+                                ] += annual_capital_cost
+
+            # 4. LINES - Capital costs (per MVA)
+            if hasattr(network, "lines") and not network.lines.empty:
+                cursor = conn.execute(
+                    """
+                    SELECT c.name as component_name, carr.name as carrier_name
+                    FROM components c
+                    JOIN carriers carr ON c.carrier_id = carr.id
+                    WHERE c.component_type = 'LINE'
+                """
+                )
+                line_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                for line_name in network.lines.index:
+                    if line_name in line_carriers:
+                        carrier_name = line_carriers[line_name]
+
+                        # Get build year and lifetime
+                        build_year = (
+                            network.lines.loc[line_name, "build_year"]
+                            if "build_year" in network.lines.columns
+                            else None
+                        )
+                        lifetime = (
+                            network.lines.loc[line_name, "lifetime"]
+                            if "lifetime" in network.lines.columns
+                            else None
+                        )
+
+                        # Check if component is active in this year
+                        if is_component_active(build_year, lifetime, year):
+                            # Get apparent power capacity and capital cost (per MVA)
+                            if "s_nom_opt" in network.lines.columns:
+                                capacity_mva = float(
+                                    network.lines.loc[line_name, "s_nom_opt"]
+                                )
+                            else:
+                                capacity_mva = (
+                                    float(network.lines.loc[line_name, "s_nom"])
+                                    if "s_nom" in network.lines.columns
+                                    else 0.0
+                                )
+
+                            capital_cost_per_mva = (
+                                float(network.lines.loc[line_name, "capital_cost"])
+                                if "capital_cost" in network.lines.columns
+                                else 0.0
+                            )
+
+                            # Calculate annualized capital cost for this year
+                            annual_capital_cost = capacity_mva * capital_cost_per_mva
+
+                            if carrier_name in carrier_stats["capital_cost_by_carrier"]:
+                                carrier_stats["capital_cost_by_carrier"][
+                                    carrier_name
+                                ] += annual_capital_cost
+
+            # 5. LINKS - Capital costs (per MW)
+            if hasattr(network, "links") and not network.links.empty:
+                cursor = conn.execute(
+                    """
+                    SELECT c.name as component_name, carr.name as carrier_name
+                    FROM components c
+                    JOIN carriers carr ON c.carrier_id = carr.id
+                    WHERE c.component_type = 'LINK'
+                """
+                )
+                link_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                for link_name in network.links.index:
+                    if link_name in link_carriers:
+                        carrier_name = link_carriers[link_name]
+
+                        # Get build year and lifetime
+                        build_year = (
+                            network.links.loc[link_name, "build_year"]
+                            if "build_year" in network.links.columns
+                            else None
+                        )
+                        lifetime = (
+                            network.links.loc[link_name, "lifetime"]
+                            if "lifetime" in network.links.columns
+                            else None
+                        )
+
+                        # Check if component is active in this year
+                        if is_component_active(build_year, lifetime, year):
+                            # Get power capacity and capital cost (per MW)
+                            if "p_nom_opt" in network.links.columns:
+                                capacity_mw = float(
+                                    network.links.loc[link_name, "p_nom_opt"]
+                                )
+                            else:
+                                capacity_mw = (
+                                    float(network.links.loc[link_name, "p_nom"])
+                                    if "p_nom" in network.links.columns
+                                    else 0.0
+                                )
+
+                            capital_cost_per_mw = (
+                                float(network.links.loc[link_name, "capital_cost"])
+                                if "capital_cost" in network.links.columns
+                                else 0.0
+                            )
+
+                            # Calculate annualized capital cost for this year
+                            annual_capital_cost = capacity_mw * capital_cost_per_mw
+
+                            if carrier_name in carrier_stats["capital_cost_by_carrier"]:
+                                carrier_stats["capital_cost_by_carrier"][
+                                    carrier_name
+                                ] += annual_capital_cost
+
+            # Calculate operational costs by carrier for this specific year
+            # Operational costs = dispatch (MWh) × marginal_cost (currency/MWh)
+            # Only for components that are active in this year
+
+            # 1. GENERATORS - Operational costs (including UNMET_LOAD)
+            if hasattr(network, "generators_t") and hasattr(network.generators_t, "p"):
+                # Get generator info: carrier, marginal_cost, build_year, lifetime (include UNMET_LOAD)
+                cursor = conn.execute(
+                    """
+                    SELECT c.name as component_name, 
+                           CASE 
+                               WHEN c.component_type = 'UNMET_LOAD' THEN 'Unmet Load'
+                               ELSE carr.name 
+                           END as carrier_name
+                    FROM components c
+                    JOIN carriers carr ON c.carrier_id = carr.id
+                    WHERE c.component_type IN ('GENERATOR', 'UNMET_LOAD')
+                """
+                )
+                generator_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                # Filter generation data for this specific year
+                year_generation = self._filter_timeseries_by_year(
+                    network.generators_t.p, network.snapshots, year
+                )
+                if year_generation is not None and not year_generation.empty:
+                    for gen_name in year_generation.columns:
+                        if gen_name in generator_carriers:
+                            carrier_name = generator_carriers[gen_name]
+
+                            # Get build year and lifetime
+                            build_year = (
+                                network.generators.loc[gen_name, "build_year"]
+                                if "build_year" in network.generators.columns
+                                else None
+                            )
+                            lifetime = (
+                                network.generators.loc[gen_name, "lifetime"]
+                                if "lifetime" in network.generators.columns
+                                else None
+                            )
+
+                            # Check if component is active in this year
+                            if is_component_active(build_year, lifetime, year):
+                                # Calculate generation for this year (already calculated above, but need to recalculate for operational costs)
+                                year_weightings = self._get_year_weightings(
+                                    network, year
+                                )
+                                if year_weightings is not None:
+                                    generation_mwh = float(
+                                        (
+                                            year_generation[gen_name].values
+                                            * year_weightings
+                                        ).sum()
+                                    )
+                                else:
+                                    generation_mwh = float(
+                                        year_generation[gen_name].sum()
+                                    )
+
+                                # Get marginal cost
+                                marginal_cost = (
+                                    float(
+                                        network.generators.loc[
+                                            gen_name, "marginal_cost"
+                                        ]
+                                    )
+                                    if "marginal_cost" in network.generators.columns
+                                    else 0.0
+                                )
+
+                                # Calculate operational cost for this year
+                                operational_cost = generation_mwh * marginal_cost
+
+                                if (
+                                    carrier_name
+                                    in carrier_stats["operational_cost_by_carrier"]
+                                ):
+                                    carrier_stats["operational_cost_by_carrier"][
+                                        carrier_name
+                                    ] += operational_cost
+
+            # 2. STORAGE_UNITS - Operational costs (discharge only)
+            if hasattr(network, "storage_units_t") and hasattr(
+                network.storage_units_t, "p"
+            ):
+                # Get storage unit info: carrier, marginal_cost, build_year, lifetime
+                cursor = conn.execute(
+                    """
+                    SELECT c.name as component_name, carr.name as carrier_name
+                    FROM components c
+                    JOIN carriers carr ON c.carrier_id = carr.id
+                    WHERE c.component_type = 'STORAGE_UNIT'
+                """
+                )
+                storage_unit_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                # Filter storage unit data for this specific year
+                year_storage = self._filter_timeseries_by_year(
+                    network.storage_units_t.p, network.snapshots, year
+                )
+                if year_storage is not None and not year_storage.empty:
+                    for su_name in year_storage.columns:
+                        if su_name in storage_unit_carriers:
+                            carrier_name = storage_unit_carriers[su_name]
+
+                            # Get build year and lifetime
+                            build_year = (
+                                network.storage_units.loc[su_name, "build_year"]
+                                if "build_year" in network.storage_units.columns
+                                else None
+                            )
+                            lifetime = (
+                                network.storage_units.loc[su_name, "lifetime"]
+                                if "lifetime" in network.storage_units.columns
+                                else None
+                            )
+
+                            # Check if component is active in this year
+                            if is_component_active(build_year, lifetime, year):
+                                # Calculate discharge for this year (positive values only)
+                                year_weightings = self._get_year_weightings(
+                                    network, year
+                                )
+                                if year_weightings is not None:
+                                    discharge_mwh = float(
+                                        (
+                                            year_storage[su_name].clip(lower=0).values
+                                            * year_weightings
+                                        ).sum()
+                                    )
+                                else:
+                                    discharge_mwh = float(
+                                        year_storage[su_name].clip(lower=0).sum()
+                                    )
+
+                                # Get marginal cost
+                                marginal_cost = (
+                                    float(
+                                        network.storage_units.loc[
+                                            su_name, "marginal_cost"
+                                        ]
+                                    )
+                                    if "marginal_cost" in network.storage_units.columns
+                                    else 0.0
+                                )
+
+                                # Calculate operational cost for this year
+                                operational_cost = discharge_mwh * marginal_cost
+
+                                if (
+                                    carrier_name
+                                    in carrier_stats["operational_cost_by_carrier"]
+                                ):
+                                    carrier_stats["operational_cost_by_carrier"][
+                                        carrier_name
+                                    ] += operational_cost
+
+            # 3. STORES - Operational costs (discharge only)
+            if hasattr(network, "stores_t") and hasattr(network.stores_t, "p"):
+                # Get store info: carrier, marginal_cost, build_year, lifetime
+                cursor = conn.execute(
+                    """
+                    SELECT c.name as component_name, carr.name as carrier_name
+                    FROM components c
+                    JOIN carriers carr ON c.carrier_id = carr.id
+                    WHERE c.component_type = 'STORE'
+                """
+                )
+                store_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                # Filter store data for this specific year
+                year_stores = self._filter_timeseries_by_year(
+                    network.stores_t.p, network.snapshots, year
+                )
+                if year_stores is not None and not year_stores.empty:
+                    for store_name in year_stores.columns:
+                        if store_name in store_carriers:
+                            carrier_name = store_carriers[store_name]
+
+                            # Get build year and lifetime
+                            build_year = (
+                                network.stores.loc[store_name, "build_year"]
+                                if "build_year" in network.stores.columns
+                                else None
+                            )
+                            lifetime = (
+                                network.stores.loc[store_name, "lifetime"]
+                                if "lifetime" in network.stores.columns
+                                else None
+                            )
+
+                            # Check if component is active in this year
+                            if is_component_active(build_year, lifetime, year):
+                                # Calculate discharge for this year (positive values only)
+                                year_weightings = self._get_year_weightings(
+                                    network, year
+                                )
+                                if year_weightings is not None:
+                                    discharge_mwh = float(
+                                        (
+                                            year_stores[store_name].clip(lower=0).values
+                                            * year_weightings
+                                        ).sum()
+                                    )
+                                else:
+                                    discharge_mwh = float(
+                                        year_stores[store_name].clip(lower=0).sum()
+                                    )
+
+                                # Get marginal cost
+                                marginal_cost = (
+                                    float(
+                                        network.stores.loc[store_name, "marginal_cost"]
+                                    )
+                                    if "marginal_cost" in network.stores.columns
+                                    else 0.0
+                                )
+
+                                # Calculate operational cost for this year
+                                operational_cost = discharge_mwh * marginal_cost
+
+                                if (
+                                    carrier_name
+                                    in carrier_stats["operational_cost_by_carrier"]
+                                ):
+                                    carrier_stats["operational_cost_by_carrier"][
+                                        carrier_name
+                                    ] += operational_cost
+
+            # Calculate total system costs by carrier for this specific year
+            # Total system cost = capital cost + operational cost
+            for carrier_name in carrier_stats["capital_cost_by_carrier"]:
+                capital_cost = carrier_stats["capital_cost_by_carrier"][carrier_name]
+                operational_cost = carrier_stats["operational_cost_by_carrier"][
+                    carrier_name
+                ]
+                total_system_cost = capital_cost + operational_cost
+
+                if carrier_name in carrier_stats["total_system_cost_by_carrier"]:
+                    carrier_stats["total_system_cost_by_carrier"][
+                        carrier_name
+                    ] = total_system_cost
+
+            # Calculate capacity by carrier for this specific year
+
+            # 4. GENERATORS - Power capacity (MW) (including UNMET_LOAD)
+            if hasattr(network, "generators") and not network.generators.empty:
+                # Get generator-carrier mapping (include UNMET_LOAD)
+                cursor = conn.execute(
+                    """
+                    SELECT c.name as component_name, 
+                           CASE 
+                               WHEN c.component_type = 'UNMET_LOAD' THEN 'Unmet Load'
+                               ELSE carr.name 
+                           END as carrier_name
+                    FROM components c
+                    JOIN carriers carr ON c.carrier_id = carr.id
+                    WHERE c.component_type IN ('GENERATOR', 'UNMET_LOAD')
+                """
+                )
+                generator_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                for gen_name in network.generators.index:
+                    if gen_name in generator_carriers:
+                        carrier_name = generator_carriers[gen_name]
+
+                        # Check if this generator is available in this year (build_year <= year)
+                        is_available = True
+                        if "build_year" in network.generators.columns:
+                            build_year = network.generators.loc[gen_name, "build_year"]
+                            if pd.notna(build_year) and int(build_year) > year:
+                                is_available = False
+
+                        if is_available:
+                            # Use p_nom_opt if available, otherwise p_nom
+                            if "p_nom_opt" in network.generators.columns:
+                                capacity_mw = float(
+                                    network.generators.loc[gen_name, "p_nom_opt"]
+                                )
+                            else:
+                                capacity_mw = (
+                                    float(network.generators.loc[gen_name, "p_nom"])
+                                    if "p_nom" in network.generators.columns
+                                    else 0.0
+                                )
+
+                            if (
+                                carrier_name
+                                in carrier_stats["power_capacity_by_carrier"]
+                            ):
+                                carrier_stats["power_capacity_by_carrier"][
+                                    carrier_name
+                                ] += capacity_mw
+
+            # 2. STORAGE_UNITS - Power capacity (MW) + Energy capacity (MWh)
+            if hasattr(network, "storage_units") and not network.storage_units.empty:
+                # Get storage unit-carrier mapping
+                cursor = conn.execute(
+                    """
+                    SELECT c.name as component_name, carr.name as carrier_name
+                    FROM components c
+                    JOIN carriers carr ON c.carrier_id = carr.id
+                    WHERE c.component_type = 'STORAGE_UNIT'
+                """
+                )
+                storage_unit_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                for su_name in network.storage_units.index:
+                    if su_name in storage_unit_carriers:
+                        carrier_name = storage_unit_carriers[su_name]
+
+                        # Check if this storage unit is available in this year
+                        is_available = True
+                        if "build_year" in network.storage_units.columns:
+                            build_year = network.storage_units.loc[
+                                su_name, "build_year"
+                            ]
+                            if pd.notna(build_year) and int(build_year) > year:
+                                is_available = False
+
+                        if is_available:
+                            # Power capacity (MW)
+                            if "p_nom_opt" in network.storage_units.columns:
+                                p_nom_opt = float(
+                                    network.storage_units.loc[su_name, "p_nom_opt"]
+                                )
+                            else:
+                                p_nom_opt = (
+                                    float(network.storage_units.loc[su_name, "p_nom"])
+                                    if "p_nom" in network.storage_units.columns
+                                    else 0.0
+                                )
+
+                            if (
+                                carrier_name
+                                in carrier_stats["power_capacity_by_carrier"]
+                            ):
+                                carrier_stats["power_capacity_by_carrier"][
+                                    carrier_name
+                                ] += p_nom_opt
+
+                            # Energy capacity (MWh) using max_hours
+                            max_hours = 1.0  # Default
+                            if "max_hours" in network.storage_units.columns:
+                                max_hours = float(
+                                    network.storage_units.loc[su_name, "max_hours"]
+                                )
+                            energy_capacity_mwh = p_nom_opt * max_hours
+
+                            if (
+                                carrier_name
+                                in carrier_stats["energy_capacity_by_carrier"]
+                            ):
+                                carrier_stats["energy_capacity_by_carrier"][
+                                    carrier_name
+                                ] += energy_capacity_mwh
+
+            # 3. STORES - Energy capacity (MWh) only
+            if hasattr(network, "stores") and not network.stores.empty:
+                # Get store-carrier mapping
+                cursor = conn.execute(
+                    """
+                    SELECT c.name as component_name, carr.name as carrier_name
+                    FROM components c
+                    JOIN carriers carr ON c.carrier_id = carr.id
+                    WHERE c.component_type = 'STORE'
+                """
+                )
+                store_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                for store_name in network.stores.index:
+                    if store_name in store_carriers:
+                        carrier_name = store_carriers[store_name]
+
+                        # Check if this store is available in this year
+                        is_available = True
+                        if "build_year" in network.stores.columns:
+                            build_year = network.stores.loc[store_name, "build_year"]
+                            if pd.notna(build_year) and int(build_year) > year:
+                                is_available = False
+
+                        if is_available:
+                            # Energy capacity (MWh)
+                            if "e_nom_opt" in network.stores.columns:
+                                capacity_mwh = float(
+                                    network.stores.loc[store_name, "e_nom_opt"]
+                                )
+                            else:
+                                capacity_mwh = (
+                                    float(network.stores.loc[store_name, "e_nom"])
+                                    if "e_nom" in network.stores.columns
+                                    else 0.0
+                                )
+
+                            if (
+                                carrier_name
+                                in carrier_stats["energy_capacity_by_carrier"]
+                            ):
+                                carrier_stats["energy_capacity_by_carrier"][
+                                    carrier_name
+                                ] += capacity_mwh
+
+            # 4. LINES - Apparent power capacity (MVA -> MW)
+            if hasattr(network, "lines") and not network.lines.empty:
+                # Get line-carrier mapping
+                cursor = conn.execute(
+                    """
+                    SELECT c.name as component_name, carr.name as carrier_name
+                    FROM components c
+                    JOIN carriers carr ON c.carrier_id = carr.id
+                    WHERE c.component_type = 'LINE'
+                """
+                )
+                line_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                for line_name in network.lines.index:
+                    if line_name in line_carriers:
+                        carrier_name = line_carriers[line_name]
+
+                        # Check if this line is available in this year
+                        is_available = True
+                        if "build_year" in network.lines.columns:
+                            build_year = network.lines.loc[line_name, "build_year"]
+                            if pd.notna(build_year) and int(build_year) > year:
+                                is_available = False
+
+                        if is_available:
+                            # Apparent power capacity (MVA -> MW, assume power factor = 1)
+                            if "s_nom_opt" in network.lines.columns:
+                                capacity_mva = float(
+                                    network.lines.loc[line_name, "s_nom_opt"]
+                                )
+                            else:
+                                capacity_mva = (
+                                    float(network.lines.loc[line_name, "s_nom"])
+                                    if "s_nom" in network.lines.columns
+                                    else 0.0
+                                )
+
+                            capacity_mw = capacity_mva  # Convert MVA to MW
+
+                            if (
+                                carrier_name
+                                in carrier_stats["power_capacity_by_carrier"]
+                            ):
+                                carrier_stats["power_capacity_by_carrier"][
+                                    carrier_name
+                                ] += capacity_mw
+
+            # 5. LINKS - Power capacity (MW)
+            if hasattr(network, "links") and not network.links.empty:
+                # Get link-carrier mapping
+                cursor = conn.execute(
+                    """
+                    SELECT c.name as component_name, carr.name as carrier_name
+                    FROM components c
+                    JOIN carriers carr ON c.carrier_id = carr.id
+                    WHERE c.component_type = 'LINK'
+                """
+                )
+                link_carriers = {row[0]: row[1] for row in cursor.fetchall()}
+
+                for link_name in network.links.index:
+                    if link_name in link_carriers:
+                        carrier_name = link_carriers[link_name]
+
+                        # Check if this link is available in this year
+                        is_available = True
+                        if "build_year" in network.links.columns:
+                            build_year = network.links.loc[link_name, "build_year"]
+                            if pd.notna(build_year) and int(build_year) > year:
+                                is_available = False
+
+                        if is_available:
+                            # Power capacity (MW)
+                            if "p_nom_opt" in network.links.columns:
+                                capacity_mw = float(
+                                    network.links.loc[link_name, "p_nom_opt"]
+                                )
+                            else:
+                                capacity_mw = (
+                                    float(network.links.loc[link_name, "p_nom"])
+                                    if "p_nom" in network.links.columns
+                                    else 0.0
+                                )
+
+                            if (
+                                carrier_name
+                                in carrier_stats["power_capacity_by_carrier"]
+                            ):
+                                carrier_stats["power_capacity_by_carrier"][
+                                    carrier_name
+                                ] += capacity_mw
+
+            return carrier_stats
+
+        except Exception as e:
+            return {
+                "dispatch_by_carrier": {},
+                "power_capacity_by_carrier": {},
+                "energy_capacity_by_carrier": {},
+                "emissions_by_carrier": {},
+                "capital_cost_by_carrier": {},
+                "operational_cost_by_carrier": {},
+                "total_system_cost_by_carrier": {},
+            }
+
+    def _sum_year_based_carrier_statistics(self, conn) -> Dict[str, Any]:
+        """
+        Sum up per-year carrier statistics for accurate multi-year totals (single network per database).
+        For capacity: take the LAST YEAR (final capacity) instead of maximum.
+        """
+        try:
+            import json
+
+            # Initialize totals
+            totals = {
+                "dispatch_by_carrier": {},
+                "power_capacity_by_carrier": {},
+                "energy_capacity_by_carrier": {},
+                "emissions_by_carrier": {},
+                "capital_cost_by_carrier": {},
+                "operational_cost_by_carrier": {},
+                "total_system_cost_by_carrier": {},
+            }
+
+            # Get all carriers from database
+            cursor = conn.execute(
+                """
+                SELECT DISTINCT name FROM carriers
+            """
+            )
+            all_carriers = [row[0] for row in cursor.fetchall()]
+
+            # Initialize all carriers with zero values (including special "Unmet Load" carrier)
+            all_carriers_with_unmet = all_carriers + ["Unmet Load"]
+            for carrier in all_carriers_with_unmet:
+                totals["dispatch_by_carrier"][carrier] = 0.0
+                totals["power_capacity_by_carrier"][carrier] = 0.0
+                totals["energy_capacity_by_carrier"][carrier] = 0.0
+                totals["emissions_by_carrier"][carrier] = 0.0
+                totals["capital_cost_by_carrier"][carrier] = 0.0
+                totals["operational_cost_by_carrier"][carrier] = 0.0
+                totals["total_system_cost_by_carrier"][carrier] = 0.0
+
+            # Get all year-based results, ordered by year
+            cursor = conn.execute(
+                """
+                SELECT year, results_json FROM network_solve_results_by_year 
+                ORDER BY year
+            """
+            )
+
+            year_results = cursor.fetchall()
+
+            if not year_results:
+                return totals
+
+            # For capacity: use the LAST YEAR only (final capacity state)
+            last_year, last_results_json = year_results[-1]
+
+            try:
+                results = json.loads(last_results_json)
+                network_stats = results.get("network_statistics", {})
+                custom_stats = network_stats.get("custom_statistics", {})
+
+                # Use last year's capacity as the all-year capacity
+                power_capacity_by_carrier = custom_stats.get(
+                    "power_capacity_by_carrier", {}
+                )
+                for carrier, value in power_capacity_by_carrier.items():
+                    if carrier in totals["power_capacity_by_carrier"]:
+                        totals["power_capacity_by_carrier"][carrier] = float(value or 0)
+
+                energy_capacity_by_carrier = custom_stats.get(
+                    "energy_capacity_by_carrier", {}
+                )
+                for carrier, value in energy_capacity_by_carrier.items():
+                    if carrier in totals["energy_capacity_by_carrier"]:
+                        totals["energy_capacity_by_carrier"][carrier] = float(
+                            value or 0
+                        )
+
+            except Exception as e:
+                pass  # Failed to process last year results
+
+            # For other stats (dispatch, emissions, costs): sum across all years
+            for year, results_json in year_results:
+                try:
+                    results = json.loads(results_json)
+                    network_stats = results.get("network_statistics", {})
+                    custom_stats = network_stats.get("custom_statistics", {})
+
+                    # Sum dispatch (energy values - sum across years)
+                    dispatch_by_carrier = custom_stats.get("dispatch_by_carrier", {})
+                    for carrier, value in dispatch_by_carrier.items():
+                        if carrier in totals["dispatch_by_carrier"]:
+                            totals["dispatch_by_carrier"][carrier] += float(value or 0)
+
+                    # Sum emissions (cumulative across years)
+                    emissions_by_carrier = custom_stats.get("emissions_by_carrier", {})
+                    for carrier, value in emissions_by_carrier.items():
+                        if carrier in totals["emissions_by_carrier"]:
+                            totals["emissions_by_carrier"][carrier] += float(value or 0)
+
+                    # Sum capital costs (cumulative across years)
+                    capital_cost_by_carrier = custom_stats.get(
+                        "capital_cost_by_carrier", {}
+                    )
+                    for carrier, value in capital_cost_by_carrier.items():
+                        if carrier in totals["capital_cost_by_carrier"]:
+                            totals["capital_cost_by_carrier"][carrier] += float(
+                                value or 0
+                            )
+
+                    # Sum operational costs (cumulative across years)
+                    operational_cost_by_carrier = custom_stats.get(
+                        "operational_cost_by_carrier", {}
+                    )
+                    for carrier, value in operational_cost_by_carrier.items():
+                        if carrier in totals["operational_cost_by_carrier"]:
+                            totals["operational_cost_by_carrier"][carrier] += float(
+                                value or 0
+                            )
+
+                    # Sum total system costs (cumulative across years)
+                    total_system_cost_by_carrier = custom_stats.get(
+                        "total_system_cost_by_carrier", {}
+                    )
+                    for carrier, value in total_system_cost_by_carrier.items():
+                        if carrier in totals["total_system_cost_by_carrier"]:
+                            totals["total_system_cost_by_carrier"][carrier] += float(
+                                value or 0
+                            )
+
+                except Exception as e:
+                    continue
+
+            return totals
+
+        except Exception as e:
+            # Return empty structure on error
+            return {
+                "dispatch_by_carrier": {},
+                "power_capacity_by_carrier": {},
+                "energy_capacity_by_carrier": {},
+                "emissions_by_carrier": {},
+                "capital_cost_by_carrier": {},
+                "operational_cost_by_carrier": {},
+                "total_system_cost_by_carrier": {},
+            }
+
+    def _serialize_results_json(self, solve_result: Dict[str, Any]) -> str:
+        """Serialize solve results to JSON string."""
+        import json
+
+        try:
+            # Create a clean results dictionary
+            results = {
+                "success": solve_result.get("success", False),
+                "status": solve_result.get("status", "unknown"),
+                "solve_time": solve_result.get("solve_time", 0.0),
+                "objective_value": solve_result.get("objective_value"),
+                "solver_name": solve_result.get("solver_name", "unknown"),
+                "run_id": solve_result.get("run_id"),
+                "network_statistics": solve_result.get("network_statistics", {}),
+                "pypsa_result": solve_result.get("pypsa_result", {}),
+            }
+            return json.dumps(results, default=self._json_serializer)
+        except Exception as e:
+            return json.dumps({"error": "serialization_failed"})
+
+    def _serialize_metadata_json(self, solve_result: Dict[str, Any]) -> str:
+        """Serialize solve metadata to JSON string."""
+        import json
+
+        try:
+            metadata = {
+                "solver_name": solve_result.get("solver_name", "unknown"),
+                "run_id": solve_result.get("run_id"),
+                "multi_period": solve_result.get("multi_period", False),
+                "years": solve_result.get("years", []),
+                "network_name": solve_result.get("network_name"),
+                "num_snapshots": solve_result.get("num_snapshots", 0),
+            }
+            return json.dumps(metadata, default=self._json_serializer)
+        except Exception as e:
+            return json.dumps({"error": "serialization_failed"})
+
+    def _filter_timeseries_by_year(
+        self, timeseries_df: "pd.DataFrame", snapshots: "pd.Index", year: int
+    ) -> "pd.DataFrame":
+        """Filter timeseries data by year"""
+        try:
+            # Handle MultiIndex case (multi-period optimization)
+            if hasattr(snapshots, "levels"):
+                period_values = snapshots.get_level_values(0)
+                year_mask = period_values == year
+                if year_mask.any():
+                    year_snapshots = snapshots[year_mask]
+                    return timeseries_df.loc[year_snapshots]
+
+            # Handle DatetimeIndex case (regular time series)
+            elif hasattr(snapshots, "year"):
+                year_mask = snapshots.year == year
+                if year_mask.any():
+                    return timeseries_df.loc[year_mask]
+
+            # Fallback - return None if can't filter
+            return None
+
+        except Exception as e:
+            return None
+
+    def _get_year_weightings(self, network: "pypsa.Network", year: int) -> "np.ndarray":
+        """Get snapshot weightings for a specific year"""
+        try:
+            # Filter snapshot weightings by year
+            if hasattr(network.snapshots, "levels"):
+                period_values = network.snapshots.get_level_values(0)
+                year_mask = period_values == year
+                if year_mask.any():
+                    year_snapshots = network.snapshots[year_mask]
+                    year_weightings = network.snapshot_weightings.loc[year_snapshots]
+                    if isinstance(year_weightings, pd.DataFrame):
+                        if "objective" in year_weightings.columns:
+                            return year_weightings["objective"].values
+                        else:
+                            return year_weightings.iloc[:, 0].values
+                    else:
+                        return year_weightings.values
+
+            elif hasattr(network.snapshots, "year"):
+                year_mask = network.snapshots.year == year
+                if year_mask.any():
+                    year_weightings = network.snapshot_weightings.loc[year_mask]
+                    if isinstance(year_weightings, pd.DataFrame):
+                        if "objective" in year_weightings.columns:
+                            return year_weightings["objective"].values
+                        else:
+                            return year_weightings.iloc[:, 0].values
+                    else:
+                        return year_weightings.values
+
+            return None
+
+        except Exception as e:
+            return None
+
+    def _calculate_total_demand(self, network: "pypsa.Network") -> float:
+        """Calculate total demand from loads in the network"""
+        try:
+            total_demand = 0.0
+
+            # Calculate demand from loads
+            if hasattr(network, "loads_t") and hasattr(network.loads_t, "p"):
+                # Apply snapshot weightings to convert MW to MWh
+                weightings = network.snapshot_weightings
+                if isinstance(weightings, pd.DataFrame):
+                    if "objective" in weightings.columns:
+                        weighting_values = weightings["objective"].values
+                    else:
+                        weighting_values = weightings.iloc[:, 0].values
+                else:
+                    weighting_values = weightings.values
+
+                total_demand = float(
+                    (network.loads_t.p.values * weighting_values[:, None]).sum()
+                )
+
+            return total_demand
+
+        except Exception as e:
+            return 0.0
+
+    def _json_serializer(self, obj):
+        """Convert numpy/pandas types to JSON serializable types"""
+        import numpy as np
+        import pandas as pd
+
+        if isinstance(obj, (np.integer, np.int64, np.int32)):
+            return int(obj)
+        elif isinstance(obj, (np.floating, np.float64, np.float32)):
+            return float(obj)
+        elif isinstance(obj, np.ndarray):
+            return obj.tolist()
+        elif isinstance(obj, pd.Series):
+            return obj.to_dict()
+        elif isinstance(obj, pd.DataFrame):
+            return obj.to_dict()
+        elif hasattr(obj, "item"):  # Handle numpy scalars
+            return obj.item()
+        else:
+            raise TypeError(f"Object of type {type(obj)} is not JSON serializable")