Pecan Street has the largest database of circuit-level, ground truth, high resolution waveform captures. If you're working on power quality, load disaggregation, grid modeling, or appliance efficiency, this is the dataset to build on.
For appliance manufacturers: Harmonic distortion has real costs downstream. Knowing that an entire appliance class injects 60-90% third harmonic current gives designers a concrete target for input filter improvement.
For NILM researchers: These consistent harmonic fingerprints are exactly what non-intrusive load monitoring algorithms need. If a refrigerator's H3/H5/H7 ratio is stable and distinctive, you can disaggregate it from whole-home power measurements.
Why does this matter?
For grid operators: understanding which devices distort the waveform and by how much becomes critical for transformer sizing and filter design.
The bottom row shows the harmonic profiles of multiple samples at each home stack almost perfectly on top of each other. Pecan Street now measures this data for all circuits in our homes (HVAC, EV chargers, washers, etc) where we see similar patterns emerge.
The top row shows raw current waveforms plotted against the 120V voltage reference. Notice how all three refrigerators produce nearly identical non-sinusoidal current shapes. The sharp peaks and flat sections are the signature of a switched-mode power supply rectifier, and they look the same.
Electrical fingerprints are consistent within and across homes.
We pulled high-frequency waveform captures from Pecan Street's database across three different home's refrigerators and ran FFT harmonic decomposition on the current draw during compressor operation.
- VPPs can help unlock off-peak grid capacity by relieving local congestion and T&D constraints
- Scaling VPPs means designing differently by sector
- Flexible demand can pair with flexible interconnection models to help connect large new datacenter loads faster
Our new blog post explores how to strategically develop, expand, and scale VPPs so we can meet projected grid demand without waiting years for new grid infrastructure. A few points worth underscoring:
Virtual power plants are underdeveloped as a grid asset yet they’re exactly the kind of quick to deploy resource the power system needs as projected load growth skyrockets.
This enables more power throughput on existing infrastructure during the vast majority of time when the grid isn’t critically congested.
Read the full report here: www.camus.energy/flexible-dat...
This is the same toolkit residential DER's offer with similar benefits to grid efficiency and growth. Pecan Street’s circuit-level data on EV charging, heat pumps, and other flexible loads shows they can shift demand during grid peaks.
A new study from Camus Energy, Princeton ZERO Lab, and Encoord shows that data centers can speed up grid interconnection by years when they combine on-site generation, batteries, and flexible computing loads.
This makes day to day load flexibility on the grid even more valuable than it is now. Residential resources like managed EV charging, home batteries, and other DERs can flex over longer intervals, not just a few peak demand hours, to help the grid operate economically and at a higher load factor.
A new Grid Strategies report shows that total electricity consumption is expected to grow faster than peak demand, shrinking the grid’s traditional downtime windows for maintenance and repairs. Read the full report here: lnkd.in/eWzvU4JA
Grid demand is projected to rise significantly in the coming years but that doesn’t just mean peak demand growth.
All of this makes customer-side resources more important than ever. Distributed solar and storage, managed EV charging, and demand response can reduce peak demand and maximize the T&D system while helping keep bills in check and improving resilience.
Historically, much of this spending was to replace aging infrastructure and harden the grid to natural disasters, but AI demand growth will likely fuel spending on T&D capacity expansion in the coming years.
Lawrence Berkeley National Lab and The Brattle Group released an excellent deep dive on what’s driving retail electricity price trends in the US. The report shows that a significant share of retail price increases stems from transmission and distribution spending.
As grid analytics and DER programs evolve, temporal precision is just as important as spatial precision.
Model validation and demand response depend on this level of visibility. AI and grid-simulation models need ground-truth signals that capture real, second-by-second load behavior instead of smoothed averages.
This chart shows residential load measured at four different resolutions. High-frequency data like Pecan Street’s 1-second circuit-level measurements reveal critical details that aren’t visible in typical smart-meter data.
Those spikes at 180 Hz, 300 Hz, etc. are electrical gremlins, harmonics that make it harder to keep voltage ‘clean’ and can shorten the life of transformers and other distribution equipment. Read more here www.pecanstreet.org/publications...
The chart below shows how much current appears at the fundamental 60 Hz frequency (what utilities want) and how much creeps in as higher frequency harmonics (what utilities don’t want).
Boo! Nothing is spookier than a little math on Halloween. We took one of our circuit level waveform captures (this one is from a washing machine) and ran a Fourier Transform to reveal its spectral skeleton.
Every home load has its own electrical signature when measured at the waveform level. We now offer ground truth, circuit level waveform data on thousands of home circuits enabling advanced research in load disaggregation, grid modeling, and power quality analysis. www.pecanstreet.org/publications...
to deploy drones + AI as an “MRI for buildings,” targeting invisible leaks, insulation gaps, and inefficiencies in real-time.
We wrote about it here → www.pecanstreet.org/2025/09/lama...
