The Electric Power Grid Potential Problems with Renewable
The Electric Power Grid: Potential Problems with Renewable Sources and New Types of Loads Michael P. Polis Oakland University Rochester, MI
Collaborators: F. Lin, C. Wang, L. Y. Wang, G. Yin, Wayne State University X. Zhang, State Grid Zhejiang Electric Power Research Institute, Hangzhou, Zhejiang B. Zhao, Southeast University, Nanjing, Jiangsu, China.
https: //www. eia. gov/energyexplained/index. php? page=electricity_delivery
Table 1. 2. A. Net Generation by Energy Source: Electric Utilities, 2017 (Thousand Megawatthours) Main Sources: Coal 2008 (1, 466, 395) 895, 095 39% Petroleum (liquid & Coke) 15, 516 Natural Gas 2008 (320, 190) 617, 725 27% Nuclear 2008 (424, 256) 424, 485 19% Hydro 2008 (229, 645) 276, 804 Solar 2008 (17) 3, 512 0. 2% Other Renewables 2008 (11, 291) 42, 618 1. 9% (wind, geothermal, wave) SUM* 2008 (2, 475, 367) 2, 270, 945 * SUM ≠ sum of Main Sources https: //www. eia. gov/electricity/monthly/epm_table_grapher. php? t=epmt_1_02_a
https: //www. osha. gov/SLTC/etools/electric_power/illustrated_glossary/transmission_lines. html
In the U. S. , the power system consists of: • more than 7, 300 power plants, • nearly 160, 000 miles of high-voltage (230500 k. V) power lines - Why high voltage? , • millions of miles of low-voltage lines(below 33 k. V & 240/120 V to the home) and distribution transformers - Why low-voltage? , It connects 145 million customers. Local electricity grids are interconnected to form larger networks for reliability and commercial purposes.
The main goal of these 3 interconnections is: To assure, in real time, that power system demand supply are balanced to avoid blackouts. The balancing authorities ensure that a sufficient supply of electricity is available to serve expected demand, which includes managing transfers of electricity with other balancing authorities. Balancing authorities are responsible for maintaining operating conditions under mandatory reliability standards.
https: //twitter. com/acore
Electric Vehicles NISSAN LEAF
Hybrid Vehicles 2018 Ford Fusion
Thus, controllable sources (as well as demands) are now appearing in the distribution network. “Smart Grid” represents a buzzword (last 10 yrs) for the many and varied approaches to dealing with the power grid that will (eventually? ) have significant penetration of renewable/controllable sources and loads in both the transmission and distribution networks.
A Microgrid A microgrid is a local energy grid which has the capability of operating independently with little or no interaction with the traditional power grid.
Work on future power systems where I’ve made some contributed • Work on energy storage systems - battery systems • Work on distributed control of future power distribution systems
Battery Systems Can the useful lifetimes of electric/hybrid vehicle batteries that must be retired as no longer useful in vehicles be extended by using them in grid storage systems? How to connect heterogeneous batteries with different characteristics and states of charge to maintain balanced State of Charge (SOC).
Battery Systems L. Y. Wang, M. P. Polis, G. G. Yin, W. Chen, Y. Fu, C. C. Mi, “Battery Cell Identification and SOC Estimation Using String Terminal Voltage Measurements, ” IEEE Trans. on Vehicular Technology, 61, 7, pp. 2925 -2935, 2012. L. Y. Wang, C. Wang, G. Yin, F. Lin, M. P. Polis, C. Zhang, and J. Jiang, Balanced Control Strategies for Interconnected Heterogeneous Battery Systems, IEEE Trans. on Sustainable Energy, 7, 1, pp. 189199, 2016 M. Liu, W. Li, C. Wang, M. P. Polis, L. Y. Wang, and J. Li, Reliability Evaluation of Large Scale Battery Energy Storage Systems, IEEE Transactions on Smart Grid, Vol, 8, Issue 6, pp. 2733 -2743, 2017.
STABILITY • Traditionally concerns generation and transmission networks. • The emergence of distributed energy generators, controllable loads, and localarea energy storage capabilities introduce potential stability concerns at the distribution network level (particularly with many PHEVs).
Distributed control of future power distribution systems L. Y. Wang, M. P. Polis, C. Wang, F. Lin, G. Yin, Voltage Robust Stability in Microgrid Power Management, Proc. 52 nd IEEE Conference on Decision and Control, Florence, Italy, Paper Fr. B 11. 1, pp. 6928 -6933, December 10 -13, 2013 B. Zhao, F. Lin, C. Wang, X. Zhang, M. P. Polis, L. Y. Wang, Supervisory Control of Networked Timed Discrete Event Systems and its Applications to Power Distribution Networks, IEEE Trans. on Control of Network Systems, Vol. 4, Issue 2, pp. 146 -158, 2017. E. Sindi, L. Y. Wang, M. P. Polis, G. Yin, L. Ding, Distributed Optimal Power and Voltage Management in DC Microgrids: Applications to Dual-Source Trolleybus Systems, submitted to IEEE Transactions on Transportation Electrification.
Voltage Robust Stability in Microgrid Power Management Proc. 52 nd IEEE Conference on Decision and Control, Florence, Italy, Paper Fr. B 11. 1, pp. 6928 -6933, December 10 -13, 2013 L. Y. Wang**, M. P. Polis*, C. Wang**, F. Lin**, G. Yin** * polis@oakland. edu * Oakland University, Rochester, MI, ** Wayne State University, Detroit, MI
This paper lays out a control-theoretic foundation for voltage stability and its robustness in microgrid systems. Load and generator dynamics are represented by node dynamic systems and their interconnections are modeled by network topologies. The combined system leads to a general nonlinear state space model.
For this model, we (1) derive stability conditions under grid constraints (2) introduce 4 different robust stability margins (3) Derive LMI-type conditions for computing the stability margins for simplified microgrids
Consider a network of microgrids shown in Fig. 1.
Let vk(t) = Vk cos (ωt + k) , so = Vk k represents the phasor voltage, and = Ik θk represents the phasor current. During transients we would have This should not be confused with the voltage time function which is a sine wave of a given frequency
For a unified state space treatment we use lower case letters to represent vectors, so :
Taking into account the number of links in the network, and the basic node current relationship at the nodes, we can write: i = Hv (4)
After some manipulation the microgrid system model becomes:
Thus, we have:
Stability Analysis We want to establish stability of the networked system: Given the load real powers, Pk and power factors, the bus voltages v 0 can be computed, and the equilibrium point x 0 can be found from
Models & Stability Conditions
We introduce four stability margins. In general, the closed form of these margins doesn’t exist, so we need to seek suitable expressions for numerical solution of each case. The goal is to express each margin in terms of linear matrix inequalities (LMI) so that efficient numerical algorithms can be used and their computational complexity can be better understood.
• For the same system as in Example 1, the total load P 1 + P 2 is plotted with different values, on top of the stability region. The maximum load that can be supported by this network is the line that is tangent to the boundary of the stability region,
We Introduce 4 Stability Margins
The 4 th Stability Margin
The 4 th Stability Margin
CONCLUSIONS This paper introduces a general framework for studying voltage stability and robustness in microgrid systems. The framework leads to a general nonlinear state space model for the entire microgrid system. Voltage stability and notions of stability margins are investigated in this framework. Although this paper uses resistive and constant-power loads as basic examples in explaining the key ideas, the framework is general and can be applied to loads of different types and more complicated link characterizations.
CONCLUSIONS There are many potential applications of this framework. For instance, distribution and load allocations of PHEV charging, battery management, and load management of microgrids can all be studied within this framework
Supervisory Control of Networked Timed Discrete Event Systems and its Applications to Power Distribution Networks, IEEE Trans. on Control of Network Systems, Vol. 4, Issue 2, pp. 146 -158, 2017. Bo Zhao 1, Feng Lin 2, Caisheng Wang 2, Xuesong Zhang 1, Michael P. Polis 3, and Le Yi Wang 2 1 State Grid Zhejiang Electric Power Research Institute, Hangzhou, Zhejiang, China, 2. Wayne State University, 3 Oakland University
Introduction • Timed Discrete Event Systems – Power distribution networks and many other systems can be modelled as discrete event systems – Time is crucial and must be considered explicitly – Discrete event systems have been investigated extensively – Some work has been done on timed discrete event systems (TDES) – No work has been done on networked TDES
Outline • • • Power Distribution Networks Timed Discrete Event Systems Networked Supervisory Control Network Controllability Network Observability Main Disadvantage of Networked Control Possible Communication delays and losses
Remarks • Network Controllability – Some events are uncontrollable – Due to delays, some controllable events may become too late to control – Due to losses, some control commands may not reach the plant – Network controllability describes the conditions under which control can still be achieved
Remarks • Network Observability – Some events are unobservable – Due to losses, some observable events may become unobservable – Due to delays, when an observable event is observed by the supervisor, some other events may have already occurred, leading the system to different states – Network observability describes the conditions under which observability can still be determined
Theoretical Results – Define network controllability – Define network observability – Prove that a networked supervisor exists if and only if both network controllability and network observability are satisfied
Power Distribution Networks
Power Distribution Networks • The supervisor is used to ensure that the total substation transformer power does not go over a pre-set critical value for too long (otherwise the circuit breaker will be tripped, causing a power outage) • Assume the supervisor can control the number of PHEVs (plug-in hybrid electric vehicles) that can be charged
Power Usage – Activity graph for substation transformer power
PHEVs being charged – Activity graph for PHEV power Ae
Controllable and enforceable events • An event is controllable if it can be disabled. • An event is enforceable if it can be enforced.
Control objectives
Power Distribution Networks • Using theoretical results, we prove that for small delays, the control objective can be achieved. • But for large delays, the control objective cannot be achieved.
Timed Discrete Event Systems
Observation delays and losses • We assume that communication delays in observation channel are random but bounded above by N. We further assume that communication is FIFO. • The observation mapping with delays and losses is denoted by
Control delays and losses • We assume that communication delays and losses in control channel are bounded by M in the following sense: delays cannot exceed M units and at least one control command in the past M time units is received by the plant. • If a control command is delayed or lost, the plant will use the most recently received control command.
Network controllability
Network observability
Existence condition
Final Remarks • If there is high penetration of distributed renewable generation and controllable loads in the power distribution system the way in which power systems are controlled will need to change. • Ignoring climate change doesn’t make the potential problems go away.
Thank you!
- Slides: 71