ANOMALY DETECTION

Your Energy Just Did WHAT?

Usage spikes are expensive surprises. We catch them in hours—not weeks—so you can fix the problem before your bill makes you cry.

The $200 Surprise Nobody Warns You About

How It Usually Goes

  1. 1. Your water heater element fails on December 1st
  2. 2. It starts running 24/7 trying to heat water
  3. 3. Your usage triples. You have no idea.
  4. 4. January 5th: $387 bill arrives
  5. 5. You call provider. "Usage was higher." Thanks.

Total damage: $200+ extra over 35 days

How It Goes With Us

  1. 1. Your water heater element fails on December 1st
  2. 2. By December 2nd, we see the spike pattern
  3. 3. You get an alert: "Usage 180% above normal"
  4. 4. We flag "24/7 flat usage = likely water heater"
  5. 5. You call a plumber. Fixed December 3rd.

Total damage: ~$15 (2 days instead of 35)

This Is What You See

Real alerts from real patterns. Not just "your usage is high"—actual intelligence about what's happening and what to do.

Single-Day Spike (Investigated)

Usage Anomaly Detected

Date

December 18, 2024

Spike Usage

98.4 kWh

Normal Usage

42.1 kWh

Cost Impact

+$8.47

Likely: HVAC Running Constantly

Ongoing Issue (Action Required)

URGENT: Multi-Day Spike

Started

December 18, 2024

Current Usage

156.2 kWh

Normal Usage

45.0 kWh

Total Cost Impact

+$52.84

Likely: Water Heater Element Failure

We Don't Just See Spikes—We Read Patterns

Different problems create different signatures. Here's how we tell them apart.

Flat & High (24/7)

Constant elevated usage with no daily pattern.

Likely: Water heater failure, stuck HVAC relay, pool pump issue

Night Spikes Only

High usage during sleeping hours, normal during day.

Likely: Space heater left on, HVAC aux heat kicking in at night

Cycling Pattern

Usage cycles on/off but runs much more than usual.

Likely: HVAC short-cycling, refrigerator compressor issue

Gradual Creep

Slow increase over days/weeks—easy to miss.

Likely: Dirty HVAC filter, aging appliance, insulation issue

Single-Day Spike

One unusual day, then back to normal.

Likely: Guests, holiday cooking, meter misread, extreme weather

Daytime Only

High usage only during business hours.

Likely: Working from home, new appliance during day, solar inverter issue

How Spike Detection Works

It's not magic. It's math, pattern recognition, and actually giving a damn.

1

Learn Your Baseline

We analyze 30+ days of your usage to understand what's normal for YOUR home. Not averages. Your actual patterns.

2

Watch for Anomalies

Our algorithms flag usage that deviates significantly from your baseline—both sudden spikes and gradual increases.

3

Alert You Fast

Get notified within hours of unusual usage—not when your bill arrives 3 weeks later.

4

Analyze the Cause

We cross-reference weather, time-of-day patterns, and your appliance profile to suggest likely causes.

5

Help You Act

Whether it's calling an HVAC tech, disputing a meter read, or just knowing to check your water heater.

The Actual Math We Use

Here's how we catch problems you'd never notice.

Baseline Calculation

// Your normal daily average
baseline = median(last_30_days)
// Account for weather
adjusted = baseline x weather_factor
// Your personal variance
threshold = adjusted + (std_dev x 2)

Spike Detection

// Today's usage vs threshold
if (today > threshold) -> ALERT
// Severity calculation
severity = (today - baseline) / baseline
// Multi-day trending
if (3_day_avg > threshold) -> URGENT

Real Example

42 kWh

Your baseline

+/-8 kWh

Normal variance

58 kWh

Alert threshold

98 kWh

Today = SPIKE

What Your Provider Says vs Reality

They have the data. They just don't share it like this.

They Say

"Your usage doubled" from provider

Reality

They tell you it happened. We tell you WHY and what to do.

They Say

"Check your meter" is their solution

Reality

We show you hourly patterns to pinpoint the exact problem.

They Say

"Meters are never wrong"

Reality

Smart meter errors exist. We help you document patterns for disputes.

They Say

"Pay now, dispute later"

Reality

We arm you with data BEFORE the bill hits.

They Say

"Probably just cold weather"

Reality

We compare your spike to your neighbors. If only you spiked, it's not weather.

When It's NOT Your Fault

Smart meters can malfunction. Reads can be estimated wrong. When the data doesn't add up, we help you build your case.

Red Flags We Help Document

  • Usage spike doesn't match neighbors during same weather

  • Spike happened while you were on vacation

  • Pattern impossible (more kWh than your main breaker allows)

  • Previous month was "estimated" and this month "caught up"

  • Usage returned to normal with no changes made

What We Provide for Disputes

  • Hourly usage data showing the anomaly

  • 30-day baseline comparison charts

  • Weather data for the period in question

  • Neighbor comparison (anonymized)

  • PDF report formatted for utility dispute submission

Important: We can't guarantee dispute outcomes—utilities make the final call. But having actual data vs "I don't think it's right" dramatically improves your chances. PUCT (Texas) regulations require utilities to respond to data-backed disputes within specific timeframes.

Questions About Spike Detection

How quickly will I know about a usage spike?

With Live Link connected, we can alert you within 4-6 hours of unusual usage starting. Without Live Link, we analyze daily data and can catch spikes within 24-48 hours. Either way, you'll know BEFORE your bill arrives.

What counts as a 'spike'?

We look for usage that's significantly higher than your baseline—typically 50%+ above your normal daily average. But we also catch 'slow leaks'—gradual 15-20% increases over several days that might indicate a failing appliance.

Can you tell exactly which appliance caused the spike?

We can't identify specific appliances without dedicated monitoring equipment, but we can analyze patterns. A spike that starts at 2am and runs 24/7? Probably water heater. Spike that correlates with outdoor temp drops? HVAC. Spike only during certain hours? Something on a schedule.

What if I think my meter is wrong?

Smart meters can malfunction—it happens more than utilities admit. We help you document the pattern, compare it to similar homes, and prepare a dispute if the data supports it. You have rights, and data is your best weapon.

Is this the same as whole-home energy monitoring?

No—dedicated monitors like Sense or Emporia give you circuit-level detail. We use smart meter data, which is less granular but doesn't require any hardware. We're great at catching spikes and patterns; circuit monitors are better at identifying specific appliances.

Can I see historical spikes?

Yes, your dashboard shows all detected anomalies with full context—when they happened, how long they lasted, what we think caused them, and how much they cost you.

Stop Getting Blindsided

The average usage spike costs $50-200 before you even know it happened. We catch them in hours, not weeks.