We have recently proposed 3-D intelligent sensors for the power grid with a view toward helping improve its reliability and restoration capability. Our solution is to use a matrix of fault tolerant distributed sensors that can sense as well as take local actions. These new sensors are "3-D Heterogeneous Sensor System on a Chip (HSoC)", which can potentially overcome delays and domino effects. The paper 1 specifically focuses on the application of the 3-D HSoCs for fault-distance estimation and the issues involved in the failure of such devices -due to defects both within the chip and due to external stress. We show that the use of multiple devices brings benefits relating to both the detection and accuracies of fault distance and arc voltage estimates. Specifically, if the nearest device is located at -x relative to the location of the transmission line fault, then the use of additional devices at -D-x, D-x, and 2D-x diminishes the probability of loss of detection considerably. The paper presents studies on this fault tolerance, as well as accuracy of estimation, vs. the number of devices used for collaborative detection. It is important to note that this collaboration does not require additional sensors, only communication among them. Matrixed HSoCs can provide several fold improvement over single HSoC.Keyword -: Smart power grid, 3-D HSoC sensors, matrixed deployment, power grid fault detection and distance estimation, defect and fault tolerance I. INTRODUCTION Electrical power grids are moving from the era of a few larger centralized power sources to one which combines these traditional systems with widely distributed green power sources. These new energy sources, such as solar and wind power, cover large geographic, often remote, areas with long transmission line connections. Due to their dependence on local environmental conditions they have considerably less predictability and control in their power generation. This combination of new challenges points to the need for the addition of smart sensors on the power grid to enhance system reliability, fault tolerance, stability and control.The underlying goal of this research is to improve the reliability and restoration capability of wide area Power Grids (PG) thereby reducing power outages and their cascade effects. Our solution is to embed novel intelligent sensors with distributed processing tools to provide real time control and monitoring of PGs from transmission and distribution levels. Our sensors will provide accurate information to protective relays, enable intelligent islanding, a means by which the modeling of loads can be based upon accurate measured data rather than estimated data for load shedding. Since power networks must be highly reliable this requires the creation of redundant and reliable systems in both the individual sensors and interaction between the sensors. The unique sensor architecture, fabrication, and signal processing issues are addressed through a "3-D Heterogeneous Sensor System on a Chip (HSoC)". While technologically and scient...