Enabling low voltage grid visibility to detect safety hazards

Wattel, Jacques Willem (2018-03)

Thesis (MEng)--Stellenbosch University, 2018.

Thesis

ENGLISH ABSTRACT: The low voltage (LV) network poses various hazards to humans and animals. In rural and deep rural areas the risk of electrocution and fire hazard is augmented, largely due to overhead conductors being used and illegal connections, with fatalities occurring each year on South Africa’s rural networks. The aim of this thesis is to investigate the possibility of detecting such hazards by employing smart meters and smart grid infrastructure. Four main hazard categories were identified as having the highest likelihood to the loss of life on Eskom’s rural LV distribution network. The hazards include: 1. PEN conductor failure. 2. LV earth electrode conductor high impedance failure. 3. Unaccounted current flow. 4. Earth leakage protection not operational. A model of each hazard is discussed and a possible detection method is proposed. An experimental setup of an LV feeder was built to verify the feasibility of the proposed detection methods. Experimental smart meters from Texas Instruments with ZigBee capabilities were installed at each customer installation and at the MV/LV transformer. The meters send measurement data to a data concentrator. The data concentrator was used to capture, store, manage and visualize data gathered from the smart meters. The thesis covers the hardware and software development that was done on the smart meters and data concentrator. Hazards were imposed on the experimental network to simulate each hazardous condition and the associated detection methods were evaluated. Lastly, the concept to monitor the feeder loop impedance using smart meters is introduced which provides an additional hazard detection parameter, providing an indication of the overall ‘health’ of a feeder. Experimental results showed that the feeder loop could successfully be measured using smart meters.

AFRIKAANSE OPSOMMING: Die laespanningnetwerk hou verskeie gevare in vir mense en diere. Die gevare is verhoog in landelike en diep landelike gebiede a.g.v hoofsaaklik oorhoofse geleiers wat gebruik word en die groot hoeveelheid onwettige konneksies. Elke jaar word verkseie sterftes vermeld a.g.v die gevare. Die hoof doel van die navorsing is om te bepaal of die gevare suksesvol bespeur kan word met die gebruik van slim meters en die slim kragnetwerk infrastruktuur. Vier hoof gevaar kategorieë was geïdentifiseer wat die hoogste gevaar inhou op Eskom se landelike laespanningnetwerke. Die gevare sluit die volgende in: 1. Breek van die PEN geleier. 2. Laespannings aardelektrode hoë impedansie fout. 3. Onverklaarde stroomvloei. 4. Aardlek beskerming buitewerking. ’n Model van elke gevaar is bespreek en ’n moontlike deteksie metode is voorgestel. ’n Experimentele opstelling van ’n laespanningvoerder was gebou om die lewensvatbaarheid van die deteksie metodes te verifieer. Experimentele slim meters van Texas Instruments met ZigBee kommunikasie was geïnstaleer by elke kliënt se installasie en by die transformator. Die meters stuur meting data na ’n data konsentrator wat die data stoor, verwerk en vertoon op ‘n webblad. Die tesis behandel die hardeware en sagteware ontwerp van die slim meters en data konsentrator. Die geïdentifiseerde gevare was geinduseer op die experimentele netwerk om die deteksie metodes te evalueer met behulp van die slim meters en data konsentrator. Laastens, is die konsep om die voerder impedansie met behulp van slim meters te bepaal, voorgestel. Dit bied ‘n addisionele gevaar deteksie parameter wat ’n goeie beraming gee van die algehele ‘gesondheid’ van ’n voerder. Eksperimentele resultate het gewys dat die voerder impedansie suksesvol bepaal kan word met die slim meters.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/103490
This item appears in the following collections: