The study examined traffic flow and safety at German motorway on‑ramps that exceed the standard design solutions defined in the German Highway Design Guidelines (RAA). In particular, double on‑ramps (EE‑type), on‑ramps with extended merge lanes, and special high‑load configurations were investigated. The research combined empirical data from permanent traffic counters and video recordings with microscopic traffic‑flow simulations performed with the BABSIM model. The simulations were used to extend the empirical data set, especially to cover conditions that could not be observed directly, such as the influence of steep gradients on heavily loaded on‑ramps.

Capacity analysis was carried out for all investigated on‑ramps by calculating the mean traffic volume before the onset of breakdown. For the EE‑1 double on‑ramps on two‑lane main roads, the capacity was found to be comparable to that of single on‑ramps of the same type, but the presence of an additional lane on the main road increased the frequency of lane‑change and sorting operations. Accident data from 44 double on‑ramps revealed that more than 60 % of incidents were caused by driver errors related to inappropriate speed, insufficient following distance, overtaking, or side‑by‑side driving. In particular, the share of side‑by‑side driving accidents on three‑lane main roads was more than twice that on two‑lane main roads. About 42 % of the accidents in categories 1–4 and 6 had a secondary cause; the most common combinations were speed as the primary cause and distance as the secondary cause on EE‑1 on‑ramps with two lanes, and distance as the primary cause on EE‑1 on‑ramps with three lanes. For EE‑2 on‑ramps, the most frequent combinations involved traffic‑ability, speed, or distance as the primary cause together with another error as the secondary cause, each accounting for 19–26 % of the incidents.

The study also evaluated a fictitious on‑ramp design that incorporated a longer merge lane and a lane‑addition on the main road. Microscopic simulation results indicated that this configuration could increase capacity by up to 10 % compared with the standard EE‑1 design while maintaining comparable safety performance. The simulations also demonstrated that the extended merge lane effectively reduced the need for abrupt lane changes, thereby lowering the likelihood of side‑by‑side collisions.

The research was conducted by the Federal Highway Research Institute (BASt) as part of its V 383 series of studies on motorway safety and traffic quality. The project was carried out over a period of several years, during which BASt collaborated with traffic‑engineering experts, simulation specialists, and accident‑analysis researchers. The project received support from the German federal government, enabling the acquisition of high‑resolution traffic‑count data and the development of advanced simulation models.

In conclusion, the findings confirm that standard on‑ramp designs remain generally safe, but high‑load situations on multi‑lane motorways benefit from advanced configurations such as double on‑ramps with lane additions or extended merge lanes. The study recommends further research into the interaction of lane‑change behaviour and traffic density on such complex junctions, as well as the integration of real‑time traffic‑management strategies to mitigate the identified accident causes.