The Hawaiian power grid is transforming from diesel to renewable.
Discover how simulation and study is helping and making engineers understand how this power grid will look in 2030.
Sandia National Laboratories uses Real-Time Simulation to Shed Light on the use of Photovoltaic Distributed Generation in Hawaii
Hawaii is heavily dependent on fossil fuel for meeting its energy needs. Indeed, more than 90% of the US state’s power is currently generated by imported foreign oil. This has left Hawaiians paying the highest energy costs in the nation, and leaves them vulnerable to foreign political instability and disruptions in supply that could cripple the economy of the Pacific island chain.
While US mainland energy costs typically hover at approximately 4% of a state’s gross domestic product (GDP), Hawaii’s costs are almost triple, approaching 11%. With the recent spike in oil prices, particularly in mid-2008 when oil prices approached US$150 per barrel, the impact on the Hawaiian economy has been dramatic, highlighted by a 36% increase in Hawaiian household fuel and utility costs during the 2nd quarter of 2008.
To address this issue, the State of Hawaii has entered into a partnership with the US Department of Energy to establish the Hawaii Clean Energy Initiative (HCEI), with the goal of having 30% of Hawaii’s energy needs met by renewable sources by 2030.
The HCEI has been launched with a focus on three projects including “Lanai 100% Renewables”. As the name suggests, the objective of this project is to assist the island of Lanai in meeting its eventual goal of obtaining 100 percent of its energy from renewable sources. The shorter term goal for Lanai is to achieve 30% of power generated from renewable sources by 2030.
Currently, Lanai is served by the Maui Electric Company, while Castle & Cooke, Inc., one of the US’s oldest real-estate developers, owns the majority of Lanai. The primary loads in Lanai come from two Castle and Cook Resorts, in addition to residential needs. The total peak load profile is 12,470V, 5.5 MW. Currently, there are several diesel generators that meet these loading requirements. As part of the HCEI, 1.2MW of Photovoltaic (PV) generation has already been installed in Lanai, bringing the HECI team very close to the 30% goal in a very short time.
PhotoVoltaic cells recently deployed in Lanai
To conduct a study evaluating the impact of integrating PV with conventional carbon-based diesel generation, the HCEI enlisted the aid of Sandia National Laboratories.
Sandia National Laboratories is a US government-owned/contractor operated facility that focuses on the development and application of technologies that ensure the homeland security of the United States. Traditionally, Sandia’s focus has been primarily placed on ensuring the safety, security and reliability of America’s nuclear weapon stockpile. However, in recent years, Sandia’s focus has increasingly turned to the development of sustainable, clean and efficient sources of energy.
Sandia engineers have faced two tasks in Lanai:
- Ensuring that the migration to renewable energy also provided for the contingency to add additional generation capacity that could then be transmitted to other islands as part of a larger state-wide power grid.
- Demonstrating that when the PV power plant is in operation, the implementation of effective controls reduces the need for capital-intensive energy storage systems for frequency and voltage stability.
For intermittent PV distributed generation, Sandia engineers investigated overall stability and transient responses. A simple Lanai “like” model was developed in the MATLAB/Simulink environment, illustrated in Figure 1, and an eMEGASim Real-Time Simulator from Opal-RT Technologies was used to conduct real-time simulation of the hybrid power grid system.
Figure 1. Lanai “like” Matlab/Simulink Power Grid Model
The diesel generators were modelled using SimPowerSystems toolbox swing equations and a custom Simulink module was developed for high-level PV generation. All of the loads were characterized primarily as distribution lines with series resistive load banks with one VAR load bank. Three-phase faults were implemented for each bus.
“The one thing we run into when adding PV to these small systems is degradation of frequency and voltage”, said Benjamin Schenkman, a member of Sandia National Laboratories’ technical team. “We’ve modelled everything in MATLAB/Simulink with the SimPowerSystems software, and so far shown found it to be stable. So, what we're trying to do is see if we need to add additional controls or add more energy storage.”
The use of simulation and Hardware-in-the-Loop testing played a critical role at this stage of the study. The non-linear power flow control models needed to be simulated to ensure that they would perform adequately now, as well as pave the way for integration of future Distributed Generation devices, such as a proposed wind farm.
“What if we add wind or more concentrated solar, how will the system react? With this project, when running tests in SimPowerSystems in offline mode, you can run maybe one test per hour. To run multiple tests takes days and days and days,” added Mr. Schenkman.
This is where the Opal-RT eMEGAsim simulator comes in. By using eMEGAsim, which is driven by RT-LAB, Opal-RT’s Real-Time Simulation platform, and ARTEMiS, an Opal-RT solver specifically designed to enable the execution of SimPowerSystems models in real-time, the simulation bottleneck was easily overcome.
According to Mr. Schenkman, “With the Opal-RT equipment, we can actually run the simulations 100 times faster, and run tests in minutes that would normally take hours. And, the results have been dead on.”
But, as is often the case with larger scale engineering projects like Lanai, nothing beats testing using physical hardware. As a consequence, Sandia conducted extensive testing at the organization’s Distributed Energy Technical Laboratory to determine whether the positive results received with the Simulink/SimPowerSystems models remained accurate.
“Customers like to see tests done with hardware, using real generators, using real PV,” said Mr. Schenkman. “So we take the Lanai grid, scale it down, take real generators, real PV. We also take the utility’s model, put it into Simulink and go to our lab. We test it out in real-time using Opal-RT equipment. And then go to Lanai and say to the customer, ‘this will work’.”
Preliminary results for generator transient responses and PV output have been positive. Both conventional and advanced control architectures are being used to evaluate the integration of the PV onto the current power grid system.
While ahead of schedule, the work in Lanai continues. Sandia engineers still face the challenge of implementing additional non-linear power flower controls into grid models and validating these models using eMEGAsim with physical Hardware-in-the-Loop at the Distributed Energy Technical Laboratory.
For both the Sandia engineers and their customer utility, the Lanai project has been a learning experience.
“The idea of this type of study is very new for utilities,” concluded Mr. Schenkman. “Using real hardware in simulations has not been their traditional approach. Utilities often just look at whether systems are balanced or unbalanced using software that is not very dynamic.”
“But by using tools like SimPowerSystems and eMEGAsim, we can remove the guesswork and can see exactly what is going to happen in the future.”
