It’s no secret rising demand for clean electricity has upended America’s utility sector, making building new fossil fuel generation a difficult proposition. But a related challenge caused by this trend is challenging utilities – the need for flexible planning and procurement to avoid risky investments.
Utilities previously accustomed to a business model of making capital investments and paying them off over decades with stable economic and policy considerations, must now build in adaptivity and frequent updates to their planning process to avoid “driving blind” on procurement. The pace of change in today’s energy markets demands it.
A new study commissioned by the U.S. Department of Energy and conducted by Lawrence Berkeley National Laboratory outlines this imperative. The study finds shifting state policies, evolving customer demand, and difficulty siting coal generation forced dramatic differences between the resources Western U.S. utilities envisioned in their Integrated Resource Plans (IRPs) and the resources they actually procured.
From 2003 to 2014, 12 utilities across the Western Interconnection added natural gas and wind instead of coal, adding three times more wind capacity than originally planned for in their IRPs . While these utility IRPs accurately forecast demand across their service territories (estimated ~20 gigawatts [GW], installed ~18 GW), their procured supplies widely diverged from forecast additions. By 2015, the utilities expected to meet new demand with 20% coal, 35% natural gas, and 15% wind, with 25% undefined – but actual procurements totaled approximately 50% wind and 50% natural gas.
Long-term planning runs into unprecedented change
IRPs evolved to expand utility planning and review processes to provide more information to regulators and better understand power supply, demand, and cost uncertainty. But renewable energy cost declines have been consistently underestimated, while policy uncertainty and siting difficulties changed in unforeseen ways between the Bush and Obama administrations, making predicting the future harder than ever for utilities.
For the 12 utilities covered in DOE’s research, long-held assumptions changed dramatically in a decade – state renewable portfolio standards required cleaner electricity supplies, long-term industrial customer demand changed, and building new coal generation became difficult because of environmental and permitting challenges. To fill this gap, the utilities bet on natural gas generation self-builds, along with wind energy power purchase agreements (PPAs) and self-builds.
The long-term trend from coal to clean isn’t news. Coal fell to 30% of America’s power mix in 2016 from 42% in 2011, typically replaced by some combination of natural gas and renewables (wind roughly doubled over this span).
The new lesson is that in an energy landscape defined by uncertainty, IRPs must account for multiple scenarios to evaluate potential changes and plan for a wide range of outcomes to meet shifting regulatory goals and customer demand. Predicting the future is nearly impossible, but the more potential market and policy changes integrated into an IRP, the more clarity it provides utilities and regulators.
Risky Bets Won’t Always Pay Off
The 12 utility IRPs largely predicted natural gas prices would fall toward the end of their forecast, and encouraged more natural gas self-builds than planned starting around 2007. But more than 80% of the new natural gas capacity deployed was built before prices actually dropped in 2009 – meaning utilities placed their bets without knowing actual prices.
In this case the utilities bet right, but building long-term generation on a hunch natural gas prices would remain low was risky at best. While fracking made natural gas cheap by the time the plants came online, historic fuel price volatility created a high degree of uncertainty and could easily convert large investments like these into stranded assets.
The utility IRPs bet wrong on renewables, but wind energy filled the gaps left by expensive coal, through far larger wind contracts and slightly greater wind self-builds than planned. Three utilities in the study (Idaho Power, PacifiCorp, and PGE) wound up with greater demand than expected, and for all three, “much larger than anticipated procurement of wind is a common trait.” Idaho and PGE procured 2-3 times more wind than planned, and PacifiCorp didn’t plan for any wind, but wind made up more than half its added generation.
As cost and policy dynamics changed for these utilities, wind energy became the easiest and often cheapest option to supply customers. Rather than locking themselves into long-term infrastructure investments that may not remain profitable years later, the utilities met demand through a mix of PPAs with wind developers or by building their own wind farms. Forward-looking utilities are moving in this direction – New Mexico utility PNM’s latest IRP announced it will stop burning coal by 2031 in favor of renewables and natural gas to lower costs, and Ohio-based AEP is building a $4.5 billion wind farm because it says wind carries far less risk than fossil fuel generation.
Utility investments based on inflexible long-term plans meet unexpected trends
The key problems utilities faced in DOE’s study are occurring across America. Once-reliable industrial customer demand forecasts are shifting as costs decline for demand-side energy management and renewable energy technology. Environmental concerns and policy uncertainty slow generation siting and contractual obligations. Rising renewable portfolio standards require more clean electricity. None of these trends were mainstream forecasts in the early 2000s.
These trends aren’t unnoticed – Utility Dive’s 2017 State of the Electric Utility Survey found utilities most confident in the growth of large-scale renewables and distributed energy resources – but according to DOE’s research, being hemmed in by long-term or inflexible IRPs fails to “create useful information for the procurement process.”
So what’s a utility executive or state regulator to do?
DOE’s research shows “most information produced in the planning phase is generally discounted from the procurement phase.” To overcome this challenge, DOE suggests utilities focus on demand forecasts and least-cost/risk portfolios during their IRP processes, and rely “extensively on the most recent information available” when making procurement decisions. In addition, DOE recommends state regulators avoid making IRPs more complex in response to technological changes and carefully define the links between IRPs and procurement.
Energy Innovation analysts recommend IRPs use (or at least analyze) the most recent publicly available data to determine cost or deployment projections for evolving technologies, require data representative of current information to triangulate projections with multiple sources, and increase flexibility through an iterative approach to improving and revising projections. IRPs should also include a range of demand and procurement forecasts with a wide range of cost assumptions under different policy settings, i.e. what mix of renewable energy would be needed to meet a high RPS target or what mix of peaking generation or transmission would be needed for reliability in a high-renewables scenario.
By combining flexible planning with current and accurate data, utilities can take advantage of the dynamic energy landscape to provide reliable electricity supply from the low-carbon generation sources customers increasingly want, without risking long-term stranded assets.
By Silvio Marcacci, Communications Director at Energy Innovation, where he leads all public relations and communications efforts.