PS10. Optimizing paving and spoils management

Underground construction unavoidably disturbs soil at the site, generating large volumes of excess spoils that must be properly processed and disposed of. Specific requirements dictate the handling of different types of spoils and result in the need for variable processing approaches depending on the moisture content, trace constituents, and size of included debris. In many cases, large volumes of spoils must be transported from remote construction sites to processing or giveaway facilities. Not only is it costly to transport spoils, but it also produces emissions from over the road trucking. The ability to cost-effectively process and dispose of spoils or reduce the overall volume generated during excavation could meaningfully reduce costs per mile across the entire undergrounding portfolio. Additionally, when undergrounding in roadways, the required excavation and subsequent paving or repaving add another. In addition to managing spoils, excavation for undergrounding in roadways typically requires extensive paving or repaving, which further increases project costs and can cause additional disruption to local communities.

PG&E has invested considerably in developing a toolkit of approaches to address the most common types of spoils generated across the undergrounding portfolio. In many cases, these efforts have focused on optimizing processes, cultivating additional partnerships for giveaway programs, and enhancing data capabilities for trackability. While these efforts have yielded cost and efficiency improvements across our spoils management efforts, there is still ample opportunity for innovative technologies to improve on-site processing capabilities, reduce the volume of spoils generated altogether, or minimize the time and expense associated with paving and road restoration.

Primary limitations include:

• Inability to avoid generation of large volumes of spoils when trenching and creation of wet spoils when boring
• Lack of on-site processing technologies that are well-suited for deployment to remote construction sites and for cost-effectively handling smaller volumes of spoils
• Lack of cost-effective techniques that minimize paving cycle times from initial dig to a fully restored road

Novel technologies, including AI/ML, to:

• Minimize the tonnage/volume of spoils from excavation that must be hauled away, by finding more cost-effective ways to process and/or repurpose spoils
• Avoid the creation of wet spoils during the boring process
• Model simulations that incorporate building-to-building transmission of fire, including radiant heat transfer and flame impingement scenarios
• Enable the cost-effective separation of sediment from wet spoils in smaller volumes than currently possible with existing technology, particularly for the required composition to be used as native backfill, while accounting for the fact that spoils must often be processed in limited space and in close proximity to residents and traffic
• Decrease the paving cycle time when undergrounding in roadways to both lower the costs of paving and to increase customer satisfaction with high quality restoration

PS11. Reducing cost and complexity of underground construction

Civil construction comprises the largest part of the total cost of undergrounding, and therefore represents the most significant opportunity for cost savings through the deployment of novel technologies. Costs may vary widely depending on several factors, including the project’s setting (urban vs. rural), terrain type, and topographical features. PG&E’s current undergrounding portfolio faces unique challenges along many of these dimensions, which make the work especially costly. Specifically, a large portion of the scheduled work is located in remote and difficult to access areas, crosses through extremely hard rock (i.e., granite), may traverse steep grades and rocky terrain or require water crossings. Addressing these challenging use cases is essential to the cost-effective completion of our 10,000-mile undergrounding program.

A variety of methods are utilized during underground construction, ranging from open cut trenching to horizontal directional drilling depending on the terrain. While specialized solutions are needed to reduce costs and improve efficiency of projects in some of our most challenging terrains, there is also a need for innovative solutions that can more efficiently complete work across broad portions of the portfolio. Specifically, there is a gap in the market for technology that enables a seamless approach to trenching, installing, and backfilling small to mid-sized conduit, as well as for boring technologies that avoid the creation of wet spoils that require processing prior to disposal. There is currently no one-size-fits-all approach or comprehensive portfolio of cost-effective point solutions that can be easily deployed across the full range of conditions encountered across the undergrounding portfolio.

Primary limitations include:

• Inability to cost effectively underground through hard rock, over steep terrain, and across canyons and water crossings
• Inability to easily maneuver in locations that are difficult to access and where space is constrained by difficult topography or existing underground facilities
• Lack of end-to-end solutions to enable the seamless burial of mid-sized conduit

Novel technologies, including AI/ML, to:

• Enable cost effective undergrounding in remote, rocky, and steep terrain, particularly with construction methods such as robotic or autonomous trenching and cable laying
• Enable efficient water crossings, compliant with all applicable regulations
• Cost-effectively construct mid-sized trenches (4-8” width; 42” depth), filling the gap between micro-trenching and other methods suited to larger conduits
• Underground runs of small electrical conduit (less than 4”) with minimal surface disruption and high navigability

PS12. Streamlining design and estimation for routine projects

Across our combined operations, PG&E completes thousands of construction projects a year, which include capacity upgrades, undergrounding projects, and routine repairs and replacements. From prioritization to design and estimation to construction, PG&E invests substantial time and resources on an annual basis completing these projects. While many of these efforts require deep expertise and creativity to solve for the specific constraints of the project, a substantial portion of this work is represented by straightforward use cases in which the design and estimation process is highly standardized from one job to the next. Novel technologies that could streamline or automate the design and estimation process for standard use cases could expand the available capacity of our estimators. This increased bandwidth could in turn be redirected to designing and estimating for complex projects, which benefit more from the expertise and creativity of our teams.

Despite a high degree of standardization across many projects, the design and estimation process relies heavily on manual effort to draft designs, specify materials, and produce estimates. This multi-step process lacks automation and requires handoffs between multiple teams, adding further complexity.

Primary limitations include:

• Lack of automated tooling or capabilities to support design and estimation of straightforward construction projects
• Increased opportunity for variability and human error due to manual nature of process
• Lack of an enterprise-wide platform to dynamically integrate pole framing into the design workflow for electric distribution projects using pole loading data, reducing manual work requirements

Novel technologies, including AI/ML, to:

• Reduce the manual effort and time to completion for designing and estimating routine projects
• Automate compliant designs using historical permitting and project data
• Increase standardization, where possible, across similar projects to reduce overall system complexity and errors in major capital projects
• Identify practical alternatives and foundational enablers that enhance design and estimation, including options that deliver reliable cost and layout insights without requiring complete 3D models
• Develop an enterprise-wide platform that automates pole framing using pole loading data and design context

PS13. Unlocking technology pathways to reduce PG&E operating costs

Affordability is foundational to customer trust, system reliability, and the success of California’s clean energy transition. PG&E’s 2027-2030 General Rate Case (GRC) application outlines a strategy to deliver customer bill stability while improving safety. PG&E forecasts no further electric rate increases in 2025, and projects that residential electric rates and average combined bills will be lower in 2026. That’s because cost recovery currently in rates will expire and be removed from rates, helping to offset proposed increases including the 2027 GRC. PG&E is requesting its smallest GRC percentage increase in a decade, made possible in part by reducing costs and passing on savings to customers. Over the past three years, the company has implemented new processes and technologies, reducing operating and capital costs by about $2.5 billion. Based on current information, if the proposal is approved in full, PG&E expects total residential combined gas and electric bills in 2027 to be flat compared to 2025 bills. Additionally, if electric demand increases—as the California Energy Commission forecasts—bills could go down because the costs of operating the system would be shared by more customers. Our goal is to stabilize bills through 2030. We are making progress and exploring every opportunity to lower energy costs and working to get the most out of every dollar spent providing safe and reliable gas and electric service to customers. Our coworkers are doing this by finding and eliminating waste in our work. We’ll also grow what’s called “beneficial load.” This means we’ll meet our customers’ higher demand for electricity (think data centers and EVs) by optimizing use of the grid, which brings down the unit price of electricity for everyone. PG&E's commitment to stabilizing bills reflects a clear operational imperative: technology must drive measurable improvements in how we operate, plan, build, and maintain the energy system. The question is no longer whether transformation is needed, but how to do so in a way that achieves sustained cost containment without compromising safety, reliability, or decarbonization goals.

While numerous efficiency initiatives are underway, PG&E’s operational cost base remains shaped by traditional methods of delivering energy services.

Primary limitations include:

• Capital- and labor-intensive field execution
• Manual operational workflows across business units
• High variability in demand forecasting and asset utilization
• Limited real-time visibility and system coordination at scale

We are looking for technology-driven solutions that can fundamentally enhance how PG&E operates across its electric business, with a focus on reducing both capital and operating expense. There is no single pathway to affordability. Rather, we are looking for:

• Step-change reductions in operational costs, effort, or time
• Scalable innovations that address persistent inefficiencies or bottlenecks (i.e., cost overruns, change orders, reverse inventory carrying costs and logistics costs)
• Systems or methods that reimagine how utility work is delivered and supported
• Tools, processes, or technologies that shift the cost curve of operations over the short- and long-term, through optimized resource utilization (labor, equipment, materials)