Historically, utilities have provided about “3 x 9s” reliability. The cost of this reliability is currently bundled into the price of electricity. This includes the cost of maintaining reserves, contingency plans, and automated generation control to cover the stochastic behavior of forced outages and electricity demand.
This cost is going up. Why?
Bulk Power Supply Uncertainty Is Affecting Reliability
The implementation of the RPS mandates is increasing the proportion of intermittent power production plants and by default decreasing the inertia, i.e., the damping ability, of the power system. As a result, a substantial amount of extra generation reserves and ancillary services are required to cover the increased uncertainty of supply while maintaining “3 x 9s” reliability levels. Recognizing this, most ISO markets trade various reserve and ancillary service products.
The transmission system is becoming more congested and there is widespread resistance against building new transmission lines. As a result, to maintain target levels of reliability and system security, more contingencies and remedial action plans and systems are needed to cover the increased uncertainty of delivery capability. Recognizing this, the ERCOT wholesale market trades month-ahead “congestion revenue right” products.
Real-World Examples of Related Supply-Side Reliability Events
A recent article by Dr. Paul-Frederik Bach, an expert in power system operations, discusses the impact of renewables penetration on the German power Grid. “The number of interventions has increased dramatically from 2010-2011 to 2011-2012…….
Bottlenecks are often detected in local grids. It makes no difference to the owner of a wind turbine if local or national grids are congested…………..In an attempt to establish an impression of the extent of interventions in Germany, EON Netz will be used as an example………..
During the first quarter of 2012, EON Netz has issued 257 interventions. The average length was 5.7 hours. Up to 10 interventions have been issued for the same hour. A total of 504 hours had one or more interventions. Thus, there have been interventions active for 23.1 percent of the hours during the first quarter of 2012..........
The total amount of curtailed energy from wind and CHP is probably modest, but the observations seem to indicate that German grids are frequently loaded to the capacity limits. Strained grids have a higher risk of cascading outages caused by single events.”
Another informative and very detailed analysis of a widespread outage in Europe in 2006 -- one which overloaded power lines and transformers in Poland by 120% and 140%, respectively -- can be found here. It includes a very interesting map of the European interconnected system showing voltage phase angle differences between substations varying from +60° to -50° across the region.
Demand-Side Uncertainty Is Also Affecting Reliability
Limited band-width, short-term frequency and voltage control is provided by traditional power plants.
However, the power industry does not have closed loop control between demand and supply. Instead, it relies on short-term forecasts of load, i.e., demand, to schedule its supply and reliability protection reserves. These forecasts are updated every few minutes. Without closed loop control, there is an unavoidable time-lag between the update of the forecast and the related adjustment of the power system to meet that demand.
Load forecasting is becoming more difficult as a result of a move towards self-optimization of electricity use by end-use customers. Power system operators are now dispatching capacity based on a forecast of the “net load” of the end-user – that is, the total usage of electricity at the customer’s site minus any DG, DR, energy storage, and automated energy management activities.
But the power system operator will not have complete visibility into the customer’s energy management actions, and often will have little or no visibility. The “net load” can exhibit unanticipated sharp and sustained up-ramps and down-ramps.
Further complicating the “net load” forecast is the increased penetration of distributed PV generation with its inherent intermittency.
As a result of all of the above, the uncertainty in the forecasts of “net load” is increasing, and thus complicating the ability of the power system operator to maintain “3 x 9s” reliability.
Aging power system infrastructure is cited as a major threat to service reliability. However, studies to date (and there are many), show that reliability measures such as SAIDI and SAIFI* appear to be only very weakly correlated with time (a proxy for the increasing age of infrastructure).
One such study appears to demonstrate an approximately 2% average annual increase in SAIDI over time, but the authors are careful to point out that the study sample is not statistically sufficient. Most published time series of SAIDI, SAIFI, CAIDI, MAIFI*, etc., as a function of time are fairly flat on a rolling average basis. This is not surprising since we have continued to observe an average “3 x 9s” service reliability for decades through to today.
An excellent, and very pragmatic, paper on the analysis and observed impacts of aging of elements of the power systems is available here.
Potential Power Distribution System Instability
There is also increasing concern about distribution system stability. Two-way flows in the distribution system are becoming more common, a regime for which distribution systems were not designed. Power injections from distributed PV generation fluctuate. EV charging occurs during off-peak periods when the power system operator may not have the flexibility to curtail some bulk power generators.
These un-planned utilization patterns have the potential to introduce instabilities into the distribution system, and negatively impact short-term reliability levels. It is not known whether and how these potential instabilities in aggregate form might affect the transmission system. To understand this better, we need to a high-technical-fidelity, dynamic, integrated simulation of transmission and distribution systems – no mean task. Some work is currently being sponsored by DOE involving computational short-cuts that might enable these types of simulations.
Real-World Examples of Distribution System Instability Challenges
Using a distribution system load flow model, Peter Evans and Soorya Kuloor showed how plug-in vehicles and distributed storage impact the grid, if not properly managed: “High penetrations of distribution connected storage devices or plug-in vehicles can have adverse grid impacts due to their charging loads.........Randomly-located or unmanaged additions, such as plug-in vehicles, can also have greater impacts at lower penetrations when compared to managed additions such as utility-sponsored storage. The studies also found that potential adverse impacts from such charging loads are highly localized, and once identified are readily managed.” It should be noted that part of this management might require the curtailment of “must-run” plants during off-peak periods for some locations.
Examples of other potential instigators of distribution system instabilities include fast-acting price-response mechanisms and pre-programmed smart appliances (using "M2M" control software), self-optimizing facilities and micro-grids, and ADR, either individually or in combinations.
Distribution load flow models have not traditionally modeled two-way load flows and intermittent distributed generation -- as a result we are still learning about distribution system state estimation under these new conditions. There are a number of efforts underway to develop low-cost micro-synchrophasors to provide the necessary real-time state estimations – see here, for example, and here.
The Costs of Lower Reliability Are Substantial for the Customer
The following widely used rules of thumb for outage costs are cited by NARUC:
- Residential: $2.50/kWh
- Commercial: $10/kWh
- Industrial: $25/kWh
Obviously, real-world numbers will be situation specific.
Meanwhile, the costs to the electricity suppliers of maintaining the current level of reliability (“3 x 9s”) are set to increase substantially.
Target-Rich Environment for Business Opportunities
It would seem that the area of reliability- and stability-related products and services provides fertile ground for entrepreneurial business opportunities.
It would also seem that the market needs new SG 2.0 applications in particular because the grid needs to be much smarter to handle the new technical challenges posed by the threats to maintaining reliability, and potentially, stability.
The SG 2.0 opportunities could include, for example, real-time control and automation systems, smart sensors, smart inverters, synchro-phasors (for both transmission and distribution systems), “big data” management and analysis, asset and/or power system optimization software, real-time state estimation, customer self-optimization software and systems, utility DG management services, ancillary services trading, renewable generation and short-term load forecasting software and systems, applications capable of reducing of SAIDI/SAIFI cash penalties for utilities, the unbundling of reliability into separate products, DA, DMS, advanced randomized DR, and perhaps an integrated catchall for end-use applications: “Smart DSM” or DSM 2.0 -- a sub-set of SG 2.0 applications -- you saw it for the first time here!
We would enjoy hearing about other opportunities that you think might become attractive businesses.
In subsequent dialogs, we will identify some of the business opportunities that we think are interesting, while expanding our discussion about the SG 2.0 applications market for reliability, stability, automation, and smart control products and services.
As always, we encourage and appreciate comments in the box below.
*Reliability: probability of satisfactory operation over the long-run
*Stability: continuance of intact operation following a disturbance
*Security: degree of risk in ability to survive imminent disturbances