Why Taiwan's Toxic Oil Maker Always Passed Its Own Safety Tests

Taiwan's edible oil industry is confronting its most serious food safety crisis in years. Roughly 1,300 metric tons of soybean salad oil produced by Central Union Oil Corp. (中聯油脂) tested positive for benzo[a]pyrene (BaP) — a Group 1 carcinogen — at four times the legal limit, sending shockwaves through the island's food supply chain.

The enforcement response, when it finally came, was forceful. The Food and Drug Administration (FDA) fined Central Union NT$160 million on July 7. Two days later, Taichung prosecutors launched coordinated raids on four linked companies across the supply chain — Central Union, Fwusow Industry (福壽), Formosa Oilseed Processing (福懋油), and Taisun Enterprise (泰山) — and sought the detention of Central Union general manager Yu Ling-chung (余凌冲) under the Act Governing Food Safety and Sanitation. After a late-night detention hearing on July 10, the Taichung District Court released Yu on NT$20 million bail; he walked out of the detention center in the early hours of July 11.

The court found grave suspicion that Yu had committed offenses including the negligent manufacture and processing of toxic food products and the use of falsified business records, and acknowledged grounds for detention existed — but ruled that because the relevant evidence had been secured, high bail would suffice.

Behind the prosecutions, however, lies a more uncomfortable story: the fatal blind spot of industry "self-management," and a bureaucracy that moved half a beat too slow.

The central absurdity of this scandal is a Rashomon-like split between the test results of Central Union and those of its customers. Namchow (南僑) first detected excessive BaP in Central Union oil through voluntary testing in May. Taisun found the same in early July. In recent days, Namchow's ongoing voluntary testing flagged abnormal readings yet again — bringing the downstream tally to three separate detections of BaP over the limit.

Central Union's own tests, meanwhile, never failed once.

The Taisun case is illustrative. When Taisun retested a batch dated May 12 and found problems, Central Union insisted its own same-day sampling had come back completely normal. Only after the FDA dispatched inspectors to examine Central Union's sampling and testing firsthand did regulators formally conclude the company's oil was defective.

The failure here is not one of laboratory technique. Central Union repeatedly stressed that its self-tests passed — but when did it draw those samples? How? Were they representative of anything? Under a self-management regime, nobody outside the company could say.

The court's bail ruling points somewhere darker. By finding grave suspicion on the falsified-records count, the court signaled that Central Union's perpetually clean self-test reports may conceal deliberate fabrication — the intentional masking of carcinogen levels above the legal limit.

Confronted with Central Union's "take our word for it" posture, the public sector's response was strikingly passive. From the scandal's full eruption on June 30 through Taisun's subsequent abnormal findings, the central regulatory apparatus sat largely idle for nearly two weeks. The turning point came on July 7, when President Lai Ching-te (賴清德) posted a sternly worded statement on Facebook. Only then did the administrative and prosecutorial machinery lurch into motion.

The first decisive act came not from Taipei but from Kaohsiung. Mayor Chen Chi-mai (陳其邁), drawing on his background in medicine and public health, judged that letting the crisis unfold batch by batch — with downstream victims surfacing one after another — would deal a devastating blow to Taiwan's food industry. Kaohsiung fired the first shot, ordering every Central Union oil product manufactured from April through June off the shelves. The Executive Yuan followed.

The FDA then went on the offensive. On July 9, the agency deployed large teams to conduct comprehensive sampling — across dates and across batch numbers — at Fwusow, Formosa Oilseed, and Central Union. The Executive Yuan had evidently concluded that the problem extended well beyond the single May 12 batch Taisun had identified; another batch from Namchow's May voluntary testing had fallen as well. According to reporting by The Journalist, the emergency samples have been sent to SGS for priority rush processing, with complete test reports covering all key April-to-June products expected within days.

The FDA's belated on-site sampling matters more than it may appear. Taiwan's two most notorious food scandals — the Ting Hsin (頂新) case and the Chang Chi (大統長基) adulterated oil case — taught Changhua prosecutors a painful lesson: in food safety litigation, if the regulator has not personally secured scientific sampling evidence, seized ledgers and verbal confessions alone will neither sustain an indictment nor secure a conviction.

A company may admit fault under the pressure of an administrative investigation. But once in court, defense counsel need only produce the company's own "all normal" test reports and argue the confession was coerced. Without the state's own laboratory evidence, judges struggle to convict. The FDA's on-site sampling arrived half a beat late — but it may prove the single decisive factor in whether the offenders can ultimately be convicted.

The scandal has also exposed a structural weakness at the top of the regulatory chain. Common sense raises an obvious question: if Central Union's process did not change and its raw materials did not change, how could only one or two batches across three months of production be contaminated?

Frontline bureaucrats hesitated to order comprehensive testing of a major supplier for a reason: risk aversion. Mass-sample a big producer, and if five of twenty samples fail, the result is a supply-chain catastrophe implicating thousands of downstream food businesses. The prevailing instinct — avoid trouble, report good news, bury the bad — meant no one dared shoulder the political responsibility. What the moment demanded was insight and nerve at the political-appointee level.

Yet Health Minister Shih Chung-liang (石崇良) and FDA Director-General Chiang Chih-kang (姜至剛) are both physicians and clinical professors by background. Clinical medicine treats individual patients; the production-and-distribution structures of the food manufacturing industry, and the operational realities of factory management, are another world entirely. That gap in industry acumen, the episode suggests, left the regulator unable to make bold early judgments — and left it trusting corporate self-management until the damage to public health and government credibility was done. （Related： Taiwan's Food Safety Crisis Is Real and Renaming It Won't Help ｜ Latest ）

Original Article in Chinese