feat: v0.7.0 — memory power draw (CPU + memory watts per CCF standard…

METHODOLOGY.md

@@ -34,16 +34,22 @@ Run `greenops-cli --coverage` to see the full instance and region list per provi
 
 ### Power Model
 
-GreenOps uses the **linear interpolation model** from the Cloud Carbon Footprint (CCF) methodology:
+GreenOps uses the **linear interpolation model** from the Cloud Carbon Footprint (CCF) methodology, extended with memory power draw:
 
 ```
-W_effective = W_idle + (W_max - W_idle) × utilization
+W_cpu    = W_idle + (W_max - W_idle) × utilization
+W_memory = memory_gb × 0.392 W/GB
+W_effective = W_cpu + W_memory
 ```
 
 Where:
 - `W_idle` = idle TDP (watts) from `factors.json`
 - `W_max` = maximum TDP (watts) from `factors.json`
 - `utilization` = CPU utilisation fraction (default: 0.50, matching CCF baseline)
+- `memory_gb` = RAM size from `factors.json`
+- `0.392 W/GB` = CCF memory power coefficient (constant, not utilization-dependent)
+
+Memory power draw is **constant** regardless of CPU utilisation. This reflects that DRAM draws near-constant power whether or not it is actively being written to, consistent with CCF v3 methodology.
 
 ### Carbon Calculation
 
@@ -64,21 +70,27 @@ GCP's 1.10 PUE is the best in class among the three major providers, producing ~
 
 ### Worked Example — AWS m5.large in us-east-1 at 50% utilisation
 
-1. **Power:** `W = 6.8 + (20.4 - 6.8) × 0.50 = 13.6W`
-2. **Energy:** `13.6W × 1.13 PUE × 730h / 1000 = 11.219 kWh/month`
-3. **Carbon:** `11.219 × 384.5 = 4,313.6g CO2e/month`
+1. **CPU power:** `W_cpu = 6.8 + (20.4 - 6.8) × 0.50 = 13.6W`
+2. **Memory power:** `W_mem = 8GB × 0.392 = 3.136W`
+3. **Total:** `W = 13.6 + 3.136 = 16.736W`
+4. **Energy:** `16.736W × 1.13 PUE × 730h / 1000 = 13.816 kWh/month`
+5. **Carbon:** `13.816 × 384.5 = 5,308.2g CO2e/month`
 
 ### Worked Example — Azure Standard_D2s_v3 in eastus at 50% utilisation
 
-1. **Power:** `W = 6.8 + (20.4 - 6.8) × 0.50 = 13.6W`
-2. **Energy:** `13.6W × 1.125 PUE × 730h / 1000 = 11.178 kWh/month`
-3. **Carbon:** `11.178 × 380.0 = 4,244.2g CO2e/month`
+1. **CPU power:** `W_cpu = 6.8 + (20.4 - 6.8) × 0.50 = 13.6W`
+2. **Memory power:** `W_mem = 8GB × 0.392 = 3.136W`
+3. **Total:** `W = 16.736W`
+4. **Energy:** `16.736W × 1.125 PUE × 730h / 1000 = 13.745 kWh/month`
+5. **Carbon:** `13.745 × 380.0 = 5,222.9g CO2e/month`
 
 ### Worked Example — GCP n2-standard-2 in us-central1 at 50% utilisation
 
-1. **Power:** `W = 6.8 + (20.4 - 6.8) × 0.50 = 13.6W`
-2. **Energy:** `13.6W × 1.10 PUE × 730h / 1000 = 10.921 kWh/month`
-3. **Carbon:** `10.921 × 340.0 = 3,713.1g CO2e/month`
+1. **CPU power:** `W_cpu = 6.8 + (20.4 - 6.8) × 0.50 = 13.6W`
+2. **Memory power:** `W_mem = 8GB × 0.392 = 3.136W`
+3. **Total:** `W = 16.736W`
+4. **Energy:** `16.736W × 1.10 PUE × 730h / 1000 = 13.445 kWh/month`
+5. **Carbon:** `13.445 × 340.0 = 4,569.3g CO2e/month`
 
 ---
 

dist/index.cjs

@@ -1968,7 +1968,7 @@ var factors_default = {
 // package.json
 var package_default = {
   name: "greenops-cli",
-  version: "0.6.0",
+  version: "0.7.0",
   description: "Carbon footprint linting for Terraform plans \u2014 AWS, Azure, and GCP. Analyses infrastructure changes for Scope 2, Scope 3, and water impact. Posts recommendations directly on GitHub PRs.",
   main: "dist/index.cjs",
   bin: {
@@ -2241,6 +2241,7 @@ function extractResourceInputs(planFilePath) {
 // engine.ts
 var HOURS_PER_MONTH = 730;
 var GRAMS_PER_KWH = 1e3;
+var MEMORY_WATTS_PER_GB = 0.392;
 function resolveUtilization(input, ledger) {
   if (input.avgUtilization !== void 0 && (input.avgUtilization < 0 || input.avgUtilization > 1)) {
     throw new RangeError(`avgUtilization must be between 0 and 1, got ${input.avgUtilization}`);
@@ -2250,8 +2251,10 @@ function resolveUtilization(input, ledger) {
   }
   return input.avgUtilization ?? ledger.metadata.assumptions.default_utilization.value;
 }
-function linearInterpolationWatts(idle, max, utilization) {
-  return idle + (max - idle) * utilization;
+function effectiveTotalWatts(idle, max, utilization, memoryGb) {
+  const cpuWatts = idle + (max - idle) * utilization;
+  const memoryWatts = memoryGb * MEMORY_WATTS_PER_GB;
+  return cpuWatts + memoryWatts;
 }
 function wattsToScope2Carbon(watts, hours, pue, gridIntensity) {
   return watts * pue * hours / GRAMS_PER_KWH * gridIntensity;
@@ -2333,7 +2336,8 @@ function calculateBaseline(input, ledger = factors_default) {
       gridIntensityApplied: gridIntensity,
       powerModelUsed: "LINEAR_INTERPOLATION",
       embodiedCo2ePerVcpuPerMonthApplied: embodied,
-      waterIntensityLitresPerKwhApplied: waterIntensity
+      waterIntensityLitresPerKwhApplied: waterIntensity,
+      memoryWattsApplied: 0
     }
   });
   const regionData = providerLedger.regions[input.region];
@@ -2359,10 +2363,11 @@ function calculateBaseline(input, ledger = factors_default) {
     );
   }
   const powerModel = "LINEAR_INTERPOLATION";
-  const effectiveWatts = linearInterpolationWatts(
+  const effectiveWatts = effectiveTotalWatts(
     instanceData.power_watts.idle,
     instanceData.power_watts.max,
-    utilization
+    utilization,
+    instanceData.memory_gb
   );
   const totalCo2eGramsPerMonth = wattsToScope2Carbon(
     effectiveWatts,
@@ -2388,7 +2393,8 @@ function calculateBaseline(input, ledger = factors_default) {
       gridIntensityApplied: regionData.grid_intensity_gco2e_per_kwh,
       powerModelUsed: powerModel,
       embodiedCo2ePerVcpuPerMonthApplied: instanceData.embodied_co2e_grams_per_month,
-      waterIntensityLitresPerKwhApplied: regionData.water_intensity_litres_per_kwh
+      waterIntensityLitresPerKwhApplied: regionData.water_intensity_litres_per_kwh,
+      memoryWattsApplied: instanceData.memory_gb * MEMORY_WATTS_PER_GB
     }
   };
 }