Residential Solar PV Systems in the Carolinas: Opportunities and Outcomes

To read the full-text of this research, you can request a copy directly from the authors.


This paper presents a first-order analysis of the feasibility and technical, environmental, and economic effects of large levels of solar Photovoltaic (PV) penetration within the services areas of the Duke Energy Carolinas (DEC) and Duke Energy Progress (DEP). A PV production model based on household density and a gridded hourly global horizontal irradiance dataset simulates hourly PV power output from roof-top installations; while a unit commitment and real time economic dispatch (UC/ED) model simulates hourly system operations. We find that the large generating capacity of base-load nuclear power plants (NPPs) without ramping capability in the region limits PV integration levels to 5.3% (6,510 MW) of 2015 generation. Enabling ramping capability for NPPs, would raise the limit of PV penetration to near 9% of electricity generated. If planned retirement of coal fired power plants together with new installations and upgrades of natural gas and nuclear plants materialize in 2025, and if NPPs operate flexibly, the share of coal-fired electricity will be reduced from 37% to 22%. A 9% penetration of electricity from PV would further reduce the share of coal-fired electricity by 4-6% resulting in a system-wide CO2 emissions rate of 0.33 tons/MWh to 0.40 tons/MWh and associated abatement costs of 225-415 (2015$/ton).

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... Third, the low operational flexibility of the several nuclear reactors operating in the service region hinders the integration of variable renewable energy [30,31]. These nuclear reactors account for 20% of the installed capacity in the DEP [32] and 25% in the DEC [33], and due to their low air emissions and marginal cost, constitute the core of the base-load supply and generate almost 50% of total electricity in the service region with capacity factors surpassing 99% [34]. ...
... In this case, the maximum hourly generation is set according to historical production information obtained from the United States Energy Information Administration (EIA) Application Program Interface [35] and the operational restrictions imposed during the systeḿs refill season as described in [36] (more information on this procedure is included in Section 5.3. of the SI). The required minimum hourly generation for units online is set as a fraction of the maximum nameplate capacity depending on the type of fuel and prime mover description as follows: 40% for coal steam turbine units [37], 30% for fuel oil combustion turbine units [38], 20% for hydroelectric units (hydraulic turbine and reversible hydraulic turbine) [38], 30% for natural gas steam turbines (both combined cycle and noncombined cycle) [39], 50% for natural gas combustion turbines (both combined cycle and non-combined cycle) [39], and 90% for nuclear reactors [6,30,31]. The generators' capability to ramp up and down their electricity production is expressed as their capacity to reach a power output equal to a fraction of their nameplate capacity in one hour. ...
... The generators' capability to ramp up and down their electricity production is expressed as their capacity to reach a power output equal to a fraction of their nameplate capacity in one hour. These capabilities are set as 100% for coal, fuel oil, hydroelectric, and natural gas [38], and 10% for nuclear reactors consistent with [6,30,31]. The minimum downtime and uptime are set based on the fuel, prime mover, and capacity as shown in Table 1 in accordance to values from previous studies [40][41][42][43][44]. ...
This study explores the performance of the Duke Energy Carolinas/Progress (DEC/DEP) electric power system under one hundred forty-one configurations combining different capacities of utility-scale photovoltaics (PV) and battery energy storage (lithium-ion batteries or BES). The different configurations include PV installations capable of providing 5–25% of the systems energy and batteries with varying duration (energy-to-power ratio) of 2, 4, and 6 h. A production cost model comprised of a day-ahead unit commitment and a real-time economic dispatch simulates the optimal operation of all the generation resources necessary to supply hourly demand and reserve requirements during the year 2016. The model represents in detail the generation fleet of the system, including 221 nuclear, natural gas, coal and hydro power generators with a combined installed capacity of 37.8 GW. Results indicate that: 1) adding BES to a power system that includes PV further reduces carbon dioxide emissions while also lowering the cost of carbon abatement. 2) The optimal power rating of a BES system that supports PV seems to be lower than 25% of the capacity of the PV. 3) BES of short duration (2-h) are more cost- effective (i.e., result in a lower cost of abatement) when the level of PV penetration is low (lower than ~12.5%), while BES of longer duration (6-h) are more cost-effective when there are larger shares of PV. 4) The installation of optimal configurations of PV +BES to reduce carbon emissions in the DEC/DEP system by ~14–57% would increase the levelized cost of electricity (LCOE) ~8–65%. 5) If projections of declining costs for the next decade materialize, the installation of up to 15 GW of PV +1.88 GW / 3.76 GWh of BES would reduce the LCOE while achieving up to 33% reduction in carbon emissions.
... China is implementing reforms [3,4] that will remove challenges for renewable energy integration that have been well known since the last decade [5], but key obstacles persist [6,7]. One of such barriers is the decentralization of scheduling and dispatch at the provincial level [8,9] and the allocation of minimum generation quotas [10,11], which result in limited ramping capability to start up and shut down conventional generators [12] as needed to balance intermittent renewable production and minimize cost and emissions. ...
... Table 3 summarizes the demand data publicly available on the websites of the 7 ISOs in the U.S. that account for more than 50% of the U.S. electricity demand. Similar data is available for regions without wholesale electricity markets (for example, the Duke Energy Carolinas/Duke Energy Progress region [12,70] or the Bonneville Power Administration region [71]. ...
The ongoing transformation of the world's energy system requires detailed power-system models that help plan a cost-effective and reliable integration of variable renewables and demand-side resources. The quality and depth of the results of these models depend on the existence of trustworthy, complete, and high-resolution data on extant electric power assets and the demand they serve, wind and solar resources, and projections on costs and performance of technologies that could be developed during the next three decades. This paper assesses the quality of China's power system's publicly available data compared to the U.S. It concludes that despite growing use of power system models to inform and analyze Chinese energy policy, the availability of necessary data is still a significant barrier that severely limits the transparency, replicability, relevance, and usefulness of their results.
... Using these results, the avoided generation cost for a year is calculated. Since the generation fixed cost data [36] considered a range for the cost: (i) high case e 0.116 $/MW-year, and (ii) low case e 0.069 $/MW-year, the avoided generation capacity cost for those two cases are calculated. Table 5 shows the results. ...
The recent increase in Renewable Generation (RG) has prompted many states, utilities, and other stakeholders to improve the methods used to determine the value of RG in efforts of replacing the Net Metering approach. However, these studies result in a wide range of values. This paper proposes a methodology based on the RG valuation studies conducted in recent years. The method includes the most common cost and benefit components considered in these studies and adopts a comprehensive method to calculate each component. The main categories include avoided energy, avoided generation capacity, avoided transmission capacity, avoided system losses, price hedging benefits, environmental benefits, and grid integration costs. A realistic case study emulating a large utility is also conducted to illustrate the application of the proposed method. The results show that all the main components can be estimated based on detailed system models or simulations. The results also illustrate some of the data challenges associated with such a study.
... economic, environmental, and social. The readers are referred to Alqahtani et al. (2016), Denholm et al. (2014), Gowrisankaran et al. (2016), Mills et al. (2013) for explorations on the other types of impacts. ...
As the number of photovoltaic (PV) installations across the world keeps on increasing, their impacts on power systems are becoming more visible and more severe. In this two-part review, the implications of high PV penetration on the stability and reliability of power systems are comprehensively assessed. This paper, the first of the two, reviews the impacts of PV on the power systems’ voltage, frequency, protection, harmonics, rotor angle stability, and flexibility requirement in detail. Factors contributing to those impacts, as well as the level and timeframe at which they occur, are carefully analysed. Subsequently, the limits of PV penetration observed in the literature are reviewed. To allow the readers to verify these impacts and limits, the tools and models typically employed in power system analysis are also elaborated. The second part of the review then completes the investigation by assessing the existing solutions to the PV integration challenges and suggesting the way forward.
... These capabilities are set as: 100% for coal, fuel oil, hydroelectric and natural gas and 10% for nuclear reactors. The minimum down time and up time are set based on the fuel, prime mover and capacity as shown in Table 2 in accordance to values from previous studies [26][27][28]. ...
Together with rising renewable energy generation, energy storage installation is also growing, both at utility scale as well as behind the metre. Behind-the-metre energy storage systems (ESS) are netted out with load and is not dispatchable by the power system operators, making them invisible. While works in the literature have investigated the benefits of coordinated ESS, the impact of non-dispatchable and invisible ESS on the transmission system remains unknown. To shed light on the topic, two levels of optimisation are formulated in this work: the upper minimises generation cost while the lower minimises customers’ electricity bill with different electricity tariffs. Through simulations at various ESS and photovoltaic (PV) penetration on IEEE reliability test system with real weather data, invisible ESS have been found to increase the power system ramping requirement, make the net demand and electricity price more volatile, as well as increase the cost of power system operations for the test system studied. In light of these findings, power system operators should analyse the impacts of invisible ESS in their respective networks to pre-emptively mitigate the possible detrimental effects.
This thesis presents a comprehensive analysis on reactive power compensation (RPC) using PV in distribution systems, which includes detailed quantification of the costs and benefits of the RPC, as well as centralised and local algorithms to optimise the reactive power dispatch of numerous PVs in the system.
This paper estimates changes in the cost of electricity, reliability, and atmospheric emissions resulting from large penetration of residential roof-top Photovoltaic (PV) and end-use energy efficiency (EE) within the service areas of Duke Energy in the Carolinas, where nuclear power plants account for almost 50% of electricity generation. Results show that 8.7–10.2% of 2015 electricity consumption could have been avoided by upgrading all residential units to comply with Energy Star standards. The range for this estimate stems from uncertainty on whether, under business-as-usual conditions, most buildings comply with the 1978 or the 1996 energy building codes. These energy savings would have implied a reduction of 3–4% in the costs of running the current power generation fleet and a 9–11% reduction in CO2 emissions. If this level of EE had been paired with the installation of roof-top PV providing 6.1–6.4% of the total electricity generated, the costs of operating the system would have been reduced by 8.6–9.6% and CO2 emissions would have been 24–26% lower. This level of roof-top PV penetration is the maximum permissible under this EE scenario due to the reductions it causes to daily peak electricity consumption and the limited operational flexibility of the nuclear plants.
Full-text available
Using six San Diego solar resource stations, clear-sky indices at 1-s resolution were computed for one site and for the average of six sites separated by less than 3 km to estimate the smoothing of aggregated power output due to geographic dispersion in a distribution feeder. Ramp rate (RR) analysis was conducted on the 1-s timeseries, including moving averages to simulate a large PV plant with energy storage. Annual maximum RRs of up to 60% per second were observed, and the largest 1-s ramp rates were enhanced over 40% by cloud reflection. However, 5% per second ramps never occurred for a simulated 10 MW power plant. Applying a wavelet transform to both the clear-sky index at one site and the average of six sites showed a strong reduction in variability at timescales shorter than 5-min, with a lesser decrease at longer timescales. Comparing these variability reductions to the Hoff and Perez (2010) model, good agreement was observed at high dispersion factors (short timescales), but our analysis shows larger reductions in variability than the model at smaller dispersion factors (long timescales).Highlights► Fluctuations in solar irradiance at various time and spatial scales were analyzed. ► Irradiance was decomposed using wavelets at timescales from 1-s to 1-h. ► Over 3 km distances ramp rates are uncorrelated at timescales less than 5 min. ► Geographic dispersion dramatically reduces aggregate short-term ramp rates.
Published data on photovoltaic (PV) degradation measurements were aggregated and re-examined. The subject has seen an increased interest in recent years resulting in more than 11 000 degradation rates in almost 200 studies from 40 different countries. As studies have grown in number and size, we found an impact from sampling bias attributable to size and accuracy. Because of the correlational nature of this study we examined the data in several ways to minimize this bias. We found median degradation for x-Si technologies in the 0.5–0.6%/year range with the mean in the 0.8–0.9%/year range. Hetero-interface technology (HIT) and microcrystalline silicon (µc-Si) technologies, although not as plentiful, exhibit degradation around 1%/year and resemble thin-film products more closely than x-Si. Several studies showing low degradation for copper indium gallium selenide (CIGS) have emerged. Higher degradation for cadmium telluride (CdTe) has been reported, but these findings could reflect a convolution of less accurate studies and longer stabilization periods for some products. Significant deviations for beginning-of-life measurements with respect to nameplate rating have been documented over the last 35 years. Therefore, degradation rates that use nameplate rating as reference may be significantly impacted. Studies that used nameplate rating as reference but used solar simulators showed less variation than similar studies using outdoor measurements, even when accounting for different climates. This could be associated with confounding effects of measurement uncertainty and soiling that take place outdoors. Hotter climates and mounting configurations that lead to sustained higher temperatures may lead to higher degradation in some, but not all, products. Wear-out non-linearities for the worst performing modules have been documented in a few select studies that took multiple measurements of an ensemble of modules during the lifetime of the system. However, the majority of these modules exhibit a fairly linear decline. Modeling these non-linearities, whether they occur at the beginning-of-life or end-of-life in the PV life cycle, has an important impact on the levelized cost of energy. Copyright
Wind integration studies are an important tool for understanding the effects of increasing wind power deployment on grid reliability and system costs. This paper provides a detailed review of the statistical methods and results from 12 large-scale regional wind integration studies. In particular, we focus our review on the modeling methods and conclusions associated with estimating short-term balancing reserves (regulation and load-following). Several important observations proceed from this review. First, we found that many of the studies either explicitly or implicitly assume that wind power step-change data follow exponential probability distributions, such as the Gaussian distribution. To understand the importance of this issue we compared empirical wind power data to Gaussian data. The results illustrate that the Gaussian assumption significantly underestimates the frequency of very large changes in wind power, and thus may lead to an underestimation of undesirable reliability effects and of operating costs. Secondly, most of these studies make extensive use of wind speed data generated from mesoscale numerical weather prediction (NWP) models. We compared the wind speed data from NWP models with empirical data and found that the NWP data have substantially less power spectral energy, a measure of variability, at higher frequencies relative to the empirical wind data. To the extent that this difference results in reduced high-frequency variability in the simulated wind power plants, studies using this approach could underestimate the need for fast ramping balancing resources. On the other hand, the magnitude of this potential underestimation is uncertain, largely because the methods used for estimating balancing reserve requirements depend on a number of heuristics, several of which are discussed in this review. Finally, we compared the power systems modeling methods used in the studies and suggest potential areas where research and development can reduce uncertainty in future wind integration studies.
A systematic framework is proposed to estimate the impact on operating costs due to uncertainty and variability in renewable resources. The framework quantifies the integration costs associated with sub-hourly variability and uncertainty as well as day-ahead forecasting errors in solar PV (photovoltaics) power. A case study illustrates how changes in system operations may affect these costs for a utility in the southwestern United States (Arizona Public Service Company). We conduct an extensive sensitivity analysis under different assumptions about balancing reserves, system flexibility, fuel prices, and forecasting errors. We find that high solar PV penetrations may lead to operational challenges, particularly during low-load and high solar periods. Increased system flexibility is essential for minimizing integration costs and maintaining reliability. In a set of sensitivity cases where such flexibility is provided, in part, by flexible operations of nuclear power plants, the estimated integration costs vary between $1.0 and $4.4/MWh-PV for a PV penetration level of 17%. The integration costs are primarily due to higher needs for hour-ahead balancing reserves to address the increased sub-hourly variability and uncertainty in the PV resource.
Adding new generation, load, or transmission to the grid changes the operation of the incumbent power system. Wind and solar generation plants are no different, but their impact on the rest of the grid is exacerbated by the facts that wind and solar energy is nondispatchable and such generators produce variable output. And because wind and solar effectively bid into the market at very low or negative cost, they are preferred resources in the dispatch stack. They are used by system operators whenever possible, unless there are generator operating limits or transmission constraints.
Due to their operational flexibility, hydroelectric dams are ideal candidates to compensate for the intermittency and unpredictability of wind energy production. However, more coordinated use of wind and hydropower resources may exacerbate the impacts dams have on downstream environmental flows, i.e., the timing and magnitude of water flows needed to sustain river ecosystems. In this paper, we examine the effects of increased (i.e., 5%, 15%, and 25%) wind market penetration on prices for electricity and reserves, and assess the potential for altered price dynamics to disrupt reservoir release schedules at a hydroelectric dam and cause more variable and unpredictable hourly flow patterns (measured in terms of the Richards-Baker Flashiness (RBF) index). Results show that the greatest potential for wind energy to impact downstream flows occurs at high (~25%) wind market penetration, when the dam sells more reserves in order to exploit spikes in real-time electricity prices caused by negative wind forecast errors. Nonetheless, compared to the initial impacts of dam construction (and the dam's subsequent operation as a peaking resource under baseline conditions) the marginal effects of any increased wind market penetration on downstream flows are found to be relatively minor.
Although solar photovoltaics (PV) are recognized as a promising source of clean energy production, researchers and policy makers need to know the optimum level of solar PV capacity penetration into the existing generation structure under the current fuel mix for the region. As the level of installed PV capacity increases, it is possible that the aggregated generation mix could produce electrical power exceeding electrical demand, thus requiring generator curtailment. Therefore, determining the optimum penetration of large-scale PV and aggregated technical and economic benefits is becoming an issue for both power utilities and policy makers. We report the development and validation of a new methodology for assessing the optimum capacity and benefits of state-wide grid-connected large scale solar PV systems in Illinois. The solar carve-out portion of the current renewable portfolio standard is also evaluated within the context of the state's sustainable energy plan for the near term future.
This report provides a full description of the Western Wind and Solar Integration Study (WWSIS) and its findings.
As photovoltaic penetration of the power grid increases, accurate predictions of return on investment require accurate prediction of decreased power output over time. Degradation rates must be known in order to predict power delivery. This article reviews degradation rates of flat-plate terrestrial modules and systems reported in published literature from field testing throughout the last 40 years. Nearly 2000 degradation rates, measured on individual modules or entire systems, have been assembled from the literature, showing a median value of 0.5%/year. The review consists of three parts: a brief historical outline, an analytical summary of degradation rates, and a detailed bibliography partitioned by technology.
Distributed generation is being deployed at increasing levels of penetration on electricity grids worldwide. It can have positive impacts on the network, but also negative impacts if integration is not properly managed. This is especially true of photovoltaics, in part because it's output fluctuates significantly and in part because it is being rapidly deployed in many countries. Potential positive impacts on grid operation can include reduced network flows and hence reduced losses and voltage drops. Potential negative impacts at high penetrations include voltage fluctuations, voltage rise and reverse power flow, power fluctuations, power factor changes, frequency regulation and harmonics, unintentional islanding, fault currents and grounding issues. This paper firstly reviews each of these impacts in detail, along with the current technical approaches available to address them. The second section of this paper discusses key non-technical factors, such as appropriate policies and institutional frameworks, which are essential to effectively coordinate the development and deployment of the different technical solutions most appropriate for particular jurisdictions. These frameworks will be different for different jurisdictions, and so no single approach will be appropriate worldwide.
In this work, we examine some of the limits to large-scale deployment of solar photovoltaics (PV) in traditional electric power systems. Specifically, we evaluate the ability of PV to provide a large fraction (up to 50%) of a utility system's energy by comparing hourly output of a simulated large PV system to the amount of electricity actually usable. The simulations use hourly recorded solar insolation and load data for Texas in the year 2000 and consider the constraints of traditional electricity generation plants to reduce output and accommodate intermittent PV generation. We find that under high penetration levels and existing grid-operation procedures and rules, the system will have excess PV generation during certain periods of the year. Several metrics are developed to examine this excess PV generation and resulting costs as a function of PV penetration at different levels of system flexibility. The limited flexibility of base load generators produces increasingly large amounts of unusable PV generation when PV provides perhaps 10–20% of a system's energy. Measures to increase PV penetration beyond this range will be discussed and quantified in a follow-up analysis.
The current practice of placing large-size fossil and nuclear steam-generating units into commercial operation will continue unabated for at least the next 10 years. Such units, intended primarily for base-load operation, generally do not have large load-turndown and two-shift cycling capabilities. Consequently, a loss of operating flexibility is slowly being created with potential future problems for power system operation.
Duke Energy Progress
Duke Energy Progress; Septmeber 2014;
PJM Renewable Integration Study; General Electric International
PJM Renewable Integration Study; General Electric International, Inc.; February 2014; 435
Integrating solar PV 448
  • J Wu
  • A Botterud
  • A Mills
  • Z Zhou
  • B Hodge
  • M Heaney
Wu, J.; Botterud, A.; Mills, A.; Zhou, Z.; Hodge, B.; and Heaney, M.; Integrating solar PV 448
  • K E Holbert
Holbert, K.E. SolarCalcs.doc. 2007; (accessed
Assessing Environmental and Reliability 503 Impacts of Flexible Ramp Products in Midcontinent ISO
  • A Cornelius
  • R Bandyopadhyay
  • D Patiño-Echeverri
Cornelius, A.; Bandyopadhyay, R.; Patiño-Echeverri, D.; Assessing Environmental and Reliability 503 Impacts of Flexible Ramp Products in Midcontinent ISO. Working paper, 2015, available at 504 505 (accessed September 30, 2015).
MacGill, I.; Watt, M.; Syngellakis, K.; The Potential Impacts of Grid
  • R Passey
  • T Spooner
Passey, R.; Spooner, T.; MacGill, I.; Watt, M.; Syngellakis, K.; The Potential Impacts of Grid- 531
Working paper, 2015; 538 sites.nicholas.duke
  • Way For
  • Thermal Solar
Way for Thermal Solar. Working paper, 2015; 538 (accessed September 30, 2015).
Duke Energy Carolinas
Integrated Resource Plan (Annual Report); Duke Energy Carolinas; September 2014; 430 (accessed 431 September 25, 2015).
For Energy RatingCD-520-33586
For Energy Rating; Report NREL/CD-520-33586; National Renewable Energy Laboratory: Golden, 469 CO, 2003; (accessed December 15, 2015).
The Potential Impacts of Grid-531
  • R Passey
  • T Spooner
  • I Macgill
  • M Watt
  • K Syngellakis
Passey, R.; Spooner, T.; MacGill, I.; Watt, M.; Syngellakis, K.; The Potential Impacts of Grid-531
  • Dc Washington
Washington, DC, 2013; 529 (accessed December 16, 2015).
Integrated Resource Plan (Annual Report. 621 622
  • Duke Energy
Duke Energy Carolinas. Integrated Resource Plan (Annual Report. 621 622 423a-9dfc-a43ec559d307 (accessed September 25, 2015). (4) 623
Power Plant Cycling Costs
  • Aptech Intertek
Intertek APTECH. Power Plant Cycling Costs, 2012. http:// 699
Integration of Renewable Resources: Technical 702
  • Iso California
California ISO. Integration of Renewable Resources: Technical 702
Wind Performance: Long -Term 669 Monitoring and Model Development For Energy Rating Report NREL/ 670 CD-520−33586
  • Photovoltaic Module
  • Thermal
Photovoltaic Module Thermal/Wind Performance: Long -Term 669 Monitoring and Model Development For Energy Rating; Report NREL/ 670 CD-520−33586; National Renewable Energy Laboratory: Golden, CO, 671 2003; (accessed De- 672 cember 15, 2015). (19) 673
Additional Costs for Load-following Nuclear PowerPlants
  • Elforsk
Elforsk. Additional Costs for Load-following Nuclear Power 721
PR-6A20−64898; National Renewable Energy 754 standards/clean-power-plan-proposed-rule
Term Projections; NREL/PR-6A20−64898; National Renewable Energy 754 standards/clean-power-plan-proposed-rule (accessed September 30, 755 2015). (41) 756
Distributed Generation System Characteristics 744
  • Icf International
ICF International. Distributed Generation System Characteristics 744
Integration of Renewable Resources: Technical Appendices for California ISO Renewable 495 Integration Studies
28. Integration of Renewable Resources: Technical Appendices for California ISO Renewable 495 Integration Studies; Caliornia ISO: Folsom, CA, 2010;
Solar Market Insight
U.S. Solar Market Insight; Solar Energy Industries Association: Washington, DC, 2015;
Integration of Renewable Resources: TechnicalAppendices for California ISO Renewable Integration Studies
  • Iso California