Kajian Teknis dan Ekonomis Penerapan Uji Emisi Kendaraan Bermotor Berbasis Mesin Pembakaran Dalam sebagai Upaya Pengendalian Pencemaran Udara di Kota Surabaya

Pertumbuhan kendaraan bermotor di Kota Surabaya yang mencapai 3,52 juta unit pada tahun 2024 dengan laju sekitar 6,0% per tahun meningkatkan urgensi pengendalian emisi gas buang dari kendaraan berbasis mesin pembakaran dalam. Penelitian ini bertujuan menganalisis karakteristik emisi karbon monoksida (CO), hidrokarbon (HC), Air Fuel Ratio (AFR), lambda, hubungan usia kendaraan dan skor perawatan terhadap emisi, efektivitas perawatan ringan dalam menurunkan emisi, perbedaan emisi antar kelompok usia kendaraan, serta kelayakan ekonomi awal program uji emisi berbasis skenario Benefit Cost Ratio (BCR) dan Net Benefit.
Penelitian menggunakan pendekatan kuantitatif empiris dengan desain deskriptif-analitik. Data utama berasal dari pengukuran emisi 30 kendaraan menggunakan gas analyzer CERTUS–CGA; analisis CO-HC-AFR/lambda difokuskan pada 28 kendaraan bensin, sedangkan 2 kendaraan diesel dianalisis terpisah melalui parameter opasitas asap. Pemilihan metode statistik mengikuti hasil uji normalitas Shapiro-Wilk: korelasi Spearman ditetapkan sebagai metode primer, uji Wilcoxon Signed-Rank digunakan untuk pengujian beda sebelum-sesudah perawatan, dan uji Kruskal-Wallis untuk perbandingan emisi antar kelompok usia. Korelasi Pearson dan regresi linear (sederhana dan berganda) digunakan sebagai metode pelengkap dengan verifikasi uji Variance Inflation Factor (VIF).
Hasil audit data kendaraan bensin menunjukkan rata-rata usia 10,96 tahun, AFR 14,614, lambda 0,994, CO sebelum perawatan 3,554% volume, CO sesudah perawatan 2,448% volume, HC sebelum perawatan 478,14 ppm, dan HC sesudah perawatan 325,93 ppm. Uji Wilcoxon menghasilkan statistik W = 0 dengan p < 0,000001 untuk penurunan CO (31,34%) dan HC (31,98%), dengan ukuran efek Cohen's d masing-masing 2,809 dan 2,970 — dikategorikan efek sangat besar. Korelasi Spearman menunjukkan skor perawatan berkorelasi sangat kuat negatif dengan CO sesudah (rₛ = −0,824; p < 0,001), sedangkan usia kendaraan berkorelasi positif kuat dengan CO sesudah (rₛ = +0,659; p < 0,001). Dalam model regresi berganda (R² = 0,692; Adj R² = 0,653; F = 17,97; p < 0,001), hanya skor perawatan yang mempertahankan signifikansi (p = 0,0003), sementara usia dan AFR tidak signifikan setelah efek skor dikontrol — menunjukkan bahwa setelah skor perawatan dimasukkan ke dalam model, pengaruh usia kendaraan tidak lagi signifikan. Hal ini mengindikasikan bahwa kondisi perawatan berperan lebih dominan dibanding usia kendaraan dalam menjelaskan variasi emisi CO pada sampel penelitian. Uji Kruskal-Wallis mengkonfirmasi perbedaan emisi CO yang signifikan antar kelompok usia (H = 8,225; p = 0,016). Nilai VIF seluruh prediktor < 3, memastikan tidak ada multikolinearitas. Analisis ekonomi moderat menghasilkan BCR 1,50 dan Net Benefit Rp100.000 per kendaraan selama tiga bulan; namun hasil ini dibatasi sebagai simulasi skenario karena konsumsi bahan bakar aktual belum diukur langsung.
Dengan demikian, uji emisi yang diikuti perawatan ringan terbukti secara statistik efektif menurunkan emisi dan berpotensi menjadi instrumen diagnosis teknis pembakaran serta alat evaluasi ekonomi awal untuk mendukung pengendalian pencemaran udara perkotaan di Kota Surabaya.

=====================================================================================

The rapid growth in the number of motor vehicles in Surabaya, which reached 3.52 million units in 2024 with an annual growth rate of approximately 6.0%, has increased the urgency of controlling exhaust emissions from internal combustion engine vehicles. This study aims to analyze the characteristics of carbon monoxide (CO), hydrocarbons (HC), air–fuel ratio (AFR), and lambda, relationship between vehicle age and maintenance score with emission levels; effectiveness of minor maintenance in reducing emissions, emission differences among vehicle age groups, and assess the preliminary economic feasibility of an emission testing program using the Benefit–Cost Ratio (BCR) and Net Benefit approaches.
This study employed an empirical quantitative approach with a descriptive-analytical research design. Primary data were obtained from emission measurements of 30 vehicles using a CERTUS–CGA gas analyzer. The analysis of CO, HC, AFR/ lambda focused on 28 gasoline-powered vehicles, while two diesel-powered vehicles were analyzed separately based on smoke opacity. Statistical methods were selected according to the results of the Shapiro–Wilk normality test. Spearman’s rank correlation was adopted as the primary correlation method, the Wilcoxon signed-rank test was used to compare emission levels before and after maintenance, and the Kruskal–Wallis test was employed to compare emissions among vehicle age groups. Pearson correlation and simple and multiple linear regression analyses were used as complementary methods, with Variance Inflation Factor (VIF) testing conducted to verify the absence of multicollinearity.
The vehicle audit showed that gasoline-powered vehicles had an average age of 10.96 years, an average AFR of 14.614, and an average lambda value of 0.994. The average CO concentration decreased from 3.554% by volume before maintenance to 2.448% by volume after maintenance, while HC concentration decreased from 478.14 ppm to 325.93 ppm. The Wilcoxon signed-rank test yielded W = 0 with p < 0.000001, indicating statistically significant reductions of 31.34% for CO and 31.98% for HC. Cohen's d effect sizes were 2.809 and 2.970, respectively, indicating very large effects. Spearman's correlation analysis revealed a very strong negative correlation between maintenance score and post- maintenance CO emissions (rₛ = −0.824, p < 0.001), whereas vehicle age showed a strong positive correlation with post- maintenance CO emissions (rₛ = +0.659, p < 0.001). In the multiple linear regression model (R² = 0.692; Adjusted R² = 0.653; F = 17.97; p < 0.001), only the maintenance score remained statistically significant (p = 0.0003), while vehicle age and AFR were no longer significant after controlling for maintenance score. These findings indicate that maintenance condition plays a more dominant role than vehicle age in explaining variations in CO emissions within the study sample. The Kruskal–Wallis test further confirmed significant differences in CO emissions among vehicle age groups (H = 8.225; p = 0.016). All predictor variables exhibited VIF values below 3, indicating the absence of multicollinearity. The moderate economic analysis yielded a Benefit–Cost Ratio (BCR) of 1.50 and a Net Benefit of IDR 100,000 per vehicle over a three-month period. However, these results should be interpreted as a scenario-based simulation because actual fuel consumption was not directly measured.
In conclusion, emission testing followed by minor maintenance has been statistically proven to effectively reduce vehicle emissions and has the potential to serve as a technical diagnostic tool for combustion performance as well as a preliminary economic evaluation instrument to support urban air pollution control in Surabaya.