The primary aim of this dissertation is to provide an efficient statistical modeling and simulation technique for extracting the dynamics of converter-dominated power systems (CDPS) with the device level details using the least amount of computation time while maintaining a reasonable level of accuracy. These test cases are intended for use by power system researchers that need to develop test cases that can perform accurate dynamic analysis on a high integration of inverter-based resources (IBRs) for stability analysis while speeding up simulation time and reducing computational complexity. Additionally, utilities/grid operators must properly plan, operate, and dispatch. Case studies that can be simulated at the device and distribution levels in power grids are presented. The dynamics of the current power grid are anticipated to change significantly with the replacement of conventional generating and the integration of distributed energy resources (DERs) as well as dynamic loads based on power electronic converters (PECs) with enhanced grid-support functions (GSFs). Given the potential advantages of such smart PECs and loads, their use is anticipated to rise. However, the dynamic behavior of the PECs and IBR-based loads while offering various ancillary services can alter dramatically with the introduction of modern control algorithms. As a result, more stochastic and nonlinear dynamics significantly challenge the stability and control of power systems. Accurate modeling of the underlying nonlinear dynamics is required to ensure the stability and dependability of the CDPS. However, the control and parameters are propriety and unknown, and as the system size increases, using a conventional modeling approach to obtain full dynamics becomes increasingly challenging and computationally expensive. Therefore, new modeling techniques are needed to accurately analyze and simulate PECs dynamics in CDPS at the device level. The first part of the work investigates existing state-of-the-art dynamic modeling techniques and tools that could embody the relatively slower electromechanical to faster electromagnetic transient (EMT) phenomena. Conventional transient stability analysis using positive-sequence simulators has become inadequate for representing CDPS, while EMT simulators suffer a high computational burden. The main goal of the first part of this research is to do a systematic review of existing dynamic simulation methods ranging from detailed switched models to computationally efficient data-driven black-box models, along with advanced co-simulation techniques, for addressing simulation speed and accuracy issues in large power system networks. Moreover, emerging hardware-based simulation tools are reviewed that reduce the computational burden and increase the simulation efficiency of the power system model. Challenges and trends in dynamic modeling and simulation in CDPS are also presented. Non-linearity is introduced in PECs dynamics as the dynamics depend on various factors, including manufacturers, physical topology, intricate models of voltage/current control loops, phase-locked loop (PLL) models, and various GSF standards. Precise design and control actions of PECs are required while providing ancillary services for accurate modeling. However, deriving detailed dynamics become challenging as the number of PECs increases, and the parameter, topology, and control strategies are proprietary and unknown, which makes modeling these PECs much more complex and computationally challenging with the available approach. The second part of this research proposes a novel method for modeling the detailed dynamics of PECs with GSFs with acceptable accuracy while consuming less computing power and speeding up the simulation. The proposed partitioned modeling algorithm utilized black-box modeling. Considering dynamic residential loads, it was tested for a smart inverter with voltage support (Volt-VAr function) on a two-bus system. The results showed a four-time speedup in simulation time compared to the use of the detailed model with acceptable levels of accuracy. The third part of this research investigates the impact of various quadrature signal generator based PLLs (QSG-PLL) methods on data-driven modeling of grid-connected single-phase inverters (GCSI) . The magnitudes of the grid voltage and current injected by a GCSI simulated in MATLAB/Simulink are estimated by each QSG-PLL approach. The best reduced-ordered dynamic partitioned model (DPM) for the GCSI model is estimated in terms of goodness of fit, using the magnitudes obtained by each approach and the instrument variable system identification (SysId) approach. Different grid disturbances (e.g., voltage sag, phase angle jumps, harmonic distortions, frequency fluctuation) are used to evaluate the performance of each data-driven model compared to the simulated model. This chapter guides researchers on which QSG-PLL to use to model GCSIs using a data-driven approach. In most previous data-driven modeling work, the SysId has been utilized to characterize the operating mode via a single DPM; nevertheless, a single linear model may be insufficient to capture crucial dynamics. The fourth part of this research proposed a partitioned modeling approach, from the second part is further extended to find the optimal number of DPM utilizing a binary search algorithm. The proposed complete simulation framework is developed that facilitates the automatic selection of DPM for training and validation, which is scalable and computationally efficient yet, accurate. The DPM was constructed using the partitioned modeling algorithm based on the input-output measurements of a PEC operating in a Volt-VAr mode that were collected in response to a more accurate probing signal. A single-house PV innovative PEC system was used to test and validate the proposed modeling approach before being tested in a modified benchmark from a Canadian suburban distribution feeder. The simulation framework created using the suggested methodology could accurately represent the dynamics of the PEC with an accuracy of more than 97% and a speedup of 2.5 times compared to the comprehensive switching model. The suggested method was modified to generate the C-code to achieve a 6.5 times speedup. Utilities could apply the proposed framework to analyze the dynamics of the power system for efficient planning, operation, and dispatch and by manufacturers to test and validate their PEC models under various operating scenarios. The utility industry has improved its CDPS models and simulations. Still, it’s not feasible to model IBR-dominated active distribution networks (DNs) with contemporary PEC-based generations and loads due to many components. This can result in incorrect data and analyses, leading to power outages. Accurate and repeatable trials require well-tested inverter models. In the final part of this research, a knowledge-free aggregated model is proposed to extend the application of the proposed partitioned modeling algorithm to obtain a general aggregated dynamic response of DERs with different topology, control actions, and internal parameters, as well as loads that are constantly changing in any DN with limited knowledge. While EMT models have made significant advancements with enhanced computing capabilities, the scalability is limited by the need for absolute knowledge about the distribution system and its components. Moreover, DER model version A (DER A) is proposed to study the aggregated dynamic behavior of DERs for stability analysis. However, the parameterization of DER A is still active research. To address these challenges, this paper proposes an aggregated model-free black-box approach for deriving a DPM of the DN. The approach incorporates dynamic loads and smart rooftop photovoltaic (PV) inverter dynamics in compliance with IEEE 1547 standards. The primary objective is to accurately represent the aggregated dynamic behavior of the CDPS for stability studies. Detailed residential distribution feeders are developed, including advanced inverter-based DERs and composite load (CMLD), from which the aggregated dynamic model is derived. The performance of the derived DPMs is evaluated through various case studies, comparing it with the detailed DN and state-of-the-art DER A model with CMLD. The data-driven DPM achieves a fitpercent of over 90%, accurately representing the aggregated dynamic behavior of DNs. Furthermore, it significantly accelerates the simulation process, up to a 68 times speedup compared to the detailed DN and 3.5 times compared to the DER A CMLD model. The secondary focus of the research is to assess the technical potential and cost-benefit analysis of the ancillary services in PEC-based load. Flexibility in PEC-based loads could be used to support the CPDS by offering a rapid, autonomous, and adjustable power reserve during system transients to help maintain system stability. Based on technical potential, ancillary service value, and implementation costs, this study illustrates the cost-benefit analysis of grid-supportive loads (GSLs) for the supply of fast frequency response (FFR). The net benefit for each GSL is demonstrated using a case study and relevant data sources. It is shown that the implementation costs for enabling GSL features are lower than the value grid operators get from acquiring responsive reserve services. Given the rising popularity of DER, GSLs can be a valuable tool for grid stability in low-inertia systems. This dissertation contributes to the power system research community by offering publicly available resources and an innovative automated modeling approach for model development and validation of PEC and CDPS. These resources enable researchers to create standardized and realistic distribution test systems to deploy their PECs. The co-simulation framework we propose can be leveraged by utilities for effective planning, operation, and dispatch and by PEC manufacturers to validate their models across different operating scenarios. ISOs and utilities implementing GSLs technology decrease the expenses of procuring FRR services while maintaining system stability.