The　distribution　network　is　at　the　end　of　the　power　system,　with　a　high　probability　of　line　faults,　among　which　single-phase　ground　faults　account　for　over　80%.　To　reduce　the　burden　of　manual　line　inspections　and　achieve　rapid　fault　restoration　in　the　distribution　network,　accurate　localization　of　single-phase　ground　faults　has　been　a　long-standing　research　focus.　However,　the　distribution　network　line　is　short,　and　the　traveling　wave　process　is　extremely　brief.　Even　with　known　fault　branch　segments,　achieving　precise　localization　of　single-phase　grounding　faults　remains　challenging.　Existing　fault　location　algorithms　often　rely　on　zero-modal　components,　but　the　instability　and　difficulty　in　estimating　the　zero-modal　wave　velocity　introduce　significant　errors　in　the　location　results.　To　address　these　issues,　this　paper　proposes　a　single-phase　grounding　fault　location　method　that　combines　the　advantages　of　traditional　single-　and　double-terminal　methods,　while　eliminating　the　dependence　on　zero-modal　wave　velocity　estimation　and　synchronization　between　terminals.　Firstly,　mathematical　calculations　of　the　on-site　modal　time　difference　are　performed　based　on　the　modal　wave　velocity　difference,　which　is　then　used　to　characterize　the　fault　location.　Variational　mode　decomposition　(VMD)　and　the　Teager　energy　operator　(TEO)　are　applied　to　separately　calibrate　the　first　wavefronts　of　the　aerial　and　zero-modal　components　on　both　sides　of　the　fault.　The　modal　time　differences　from　both　sides　are　then　substituted　into　the　fault　location　formula　to　calculate　the　fault　distance.　This　approach　eliminates　the　need　for　estimating　the　zero-modal　wave　velocity　and　avoids　the　challenges　of　synchronization　and　detecting　reflected　waves,　effectively　improving　the　accuracy　of　fault　location.　Simulation　results　based　on　PSCAD/EMTDC　show　that　the　proposed　fault　location　method　can　quickly　and　reliably　locate　single-phase　grounding　faults　in　overhead　distribution　network　lines,　with　an　error　of　less　than　100　meters　across　various　fault　scenarios,　including　different　fault　resistances,　fault　angles,　fault　locations,　fault　phases,　and　grounding　methods　of　the　neutral　point.　Testing　has　shown　that　the　noise　resistance　threshold　of　the　proposed　algorithm　is　40　dB.　When　the　noise　level　in　the　measurement　environment　is　high　(SNR　＜　40　dB)　and　there　is　a　certain　proportion　of　impulse　noise,　the　combination　of　multiple　moving　average　filters　(to　suppress　white　noise)　and　median　filters　(to　suppress　impulse　noise)　ensures　the　accuracy　of　wavefront　calibration.　Results　from　dynamic　simulation　tests　demonstrate　that　the　proposed　method　can　still　accurately　locate　faults　in　high-noise　environments,　with　a　relative　error　of　1.26%.　Finally,　based　on　numerical　analysis　theory,　error　bounds　are　derived　for　both　the　single-　and　double-terminal　methods　based　on　the　modulus　wave　speed　difference,　as　well　as　for　the　proposed　algorithm.　The　impact　of　wavefront　calibration　errors　and　zero-modal　wave　velocity　estimation　errors　on　the　error　bounds　is　explored,　revealing　the　theoretical　basis　for　the　improved　accuracy.　Validation　cases　using　the　PSCAD/EMTDC　simulation　platform　confirm　these　findings.　The　results　indicate　that,　without　considering　synchronization　errors,　the　proposed　method　and　the　double-terminal　method　achieve　significantly　higher　accuracy　than　the　algorithm　based　on　the　modal　wave　velocity　difference.　After　accounting　for　synchronization　errors,　the　proposed　method　demonstrates　higher　accuracy　than　the　double-terminal　method.　©　2025　China　Machine　Press.　All　rights　reserved.