York Oil Company (“YOCO”) appeals from an order dated 11 February 2003 affirming a final agency decision dated 11 February 2000 by the North Carolina Department of Environment, Health and Natural Resources (“NCDEHNR”) denying YOCO eligibility to receive reimbursement for clean up costs from the Commercial Leaking Petroleum Underground Storage Tank Cleanup Fund, N.C. Gen. Stat. § 143-215.94B (2003), (“the Fund”). Because summary judgment was improperly granted, we reverse and remand.
The evidence contained in the record on appeal tends to show the following. YOCO has owned underground storage tanks (“USTs”) located at the One-Stop gas station (“One-Stop”) on Vance Road in Kernersville, North Carolina, since 1979 and installed new USTs in 1981. See James v. Clark, 118 N.C. App. 178, 179-80, 454 S.E.2d 826, 827 (1995). One-Stop is owned by David Clark (“Clark”). In July 1986, Walter James (“James”) who owned property neighboring One-Stop reported to the regional office of NCDEHNR that his well was contaminated with gasoline. Subsequent investigation of James’ complaint by Stephen Kay, an NCDEHNR employee, revealed that the water on James’ property had been contaminated for about five years and smelled heavily of gasoline. One-Stop was the only gas station within *552a half-mile radius of James’ property and the James’ well was located 150 feet down gradient from the USTs.
Kay interviewed both the manager of One-Stop and Clark, the owner. The store manager stated that the store bought bottled water for drinking and that a gasoline odor could sometimes be detected when the toilets were flushed. Clark informed Kay that contamination in the water from One-Stop’s own well had been noticeable since one or two years after the well’s installation in 1982. Clark did not recall a conversation about contamination, but did recall a conversation about the septic system. Kay concluded in his report that One-Stop was the only possible source of the contamination and arranged for monitoring wells to be placed nearby to establish the extent of the contamination as well as to gather evidence to support a notice of violation.
As a result of the reports of contamination on the neighboring property, YOCO hired Collins Petroleum to perform some testing. In a letter not dated until 23 August 1988, Collins Petroleum stated that it had dug eighteen inches below the bottom of two of the USTs to look for leaks and had found none, but had discovered the odor of gasoline above the tanks. In September 1986, a letter was sent by the Forsyth County Health Department to Clark informing him that test results showed One-Stop’s water supply tested positive for fecal col-iform bacteria and in addition petroleum contamination was suspected at One-Stop and that further testing was being done. A 12 September 1986 newspaper article in a local paper revealed that NCDEHNR had in fact discovered the James’ water to be contaminated with gasoline probably from leaking USTs. Although denying he ever received official notification of the testing, Clark acknowledged that he had read the newspaper article and had given it to YOCO. In a subsequent deposition, Gary York, the owner of YOCO, admitted that someone had made him aware of a problem with contamination or spillage of petroleum on an adjoining property in 1986. An analysis of a water sample taken from One-Stop in November 1986, however, revealed that there was “[n]o base/neutral or acid extractable organics detected.”
On 11 March 1988, two and a half feet of gasoline was discovered in a monitoring well located at One-Stop. A letter dated 20 May 1988 addressed to James indicated that NCDEHNR had not made any determinations from its investigation of the contamination of James’ property. On 28 November 1988, NCDEHNR issued a draft report concluding the contamination of James’ water supply was caused by *553leaking USTs at One-Stop. As a result of the March 1988 discovery, a notice of violation was ultimately sent to YOCO and Clark on 10 February 1989.
On 17 April 1997, YOCO applied for reimbursement from the Fund for expenses related to the clean up of leaking UST’s. The application was denied by NCDEHNR on 17 June 1997 because the leakage had been “discovered” prior to the Fund’s effective date of 30 June 1988. See N.C. Gen. Stat. § 143-215.94N (2003). On 14 August 1997, YOCO filed a petition for a contested case hearing arguing that YOCO had not been made aware of the leak until 1989. On 15 April 1999, both NCDEHNR and YOCO moved for summary judgment before the Administrative Law Judge (“AU”). The AU in a recommended decision concluded that the denial of YOCO’s eligibility to receive reimbursement from the fund was proper and granted summary judgment for NCDEHNR. In a final agency decision dated 11 February 2000, NCDEHNR adopted the recommended decision of the ALJ and affirmed the denial of reimbursement under the Fund.
YOCO petitioned for judicial review of the decision before the trial court. YOCO also sought to have the trial court consider a letter issued by NCDEHNR on 2 April 2001 in a separate matter, which indicated that a single report of odor of gasoline alone was insufficient to support a conclusion that a leak had been detected prior to the effective date of the Fund in determining eligibility to receive reimbursement. The trial court refused to consider this letter as it was not part of the record submitted from the final agency decision. In its 11 February 2003 decision, the trial court affirmed the final agency decision.
The dispositive issue is whether the trial court properly affirmed the final agency decision adopting summary judgment in favor of NCDEHNR.1 Specifically, YOCO contends that (A) in granting summary judgment in favor of NCDEHNR, the ALJ applied the wrong legal standard as to whether YOCO had discovered the release prior to the effective date of the Fund, and (B) there was a genuine issue of material fact as to whether the release had been discovered prior to the effective date of the Fund.
*554“In reviewing a final agency decision allowing .. . summary judgment . . . , the [trial] court may enter any order allowed by . . . Rule 56.” N.C. Gen. Stat. § 150B-51(d) (2003). The role of an appellate court in reviewing a trial court’s order affirming a decision by an administrative agency is two-fold. In re Appeal by McCrary, 112 N.C. App. 161, 166, 435 S.E.2d 359, 363 (1993). We must: “(1) determine the appropriate standard of review and, when applicable, (2) determine whether the trial court properly applied this standard.” Id. De novo review is applied where an error of law is alleged. See id. “When the issue on appeal is whether a state agency erred in interpreting a regulatory term, an appellate court may freely substitute its judgment for that of the agency and employ de novo review.” Britt v. N.C. Sheriffs’ Educ. and Training Stds. Comm’n, 348 N.C. 573, 576, 501 S.E.2d 75, 77 (1998). In addition, the grant of summary judgment involves a matter of law, which is reviewable de novo. See Falk Integrated Tech., Inc. v. Stack, 132 N.C. App. 807, 809, 513 S.E.2d 572, 574 (1999).
A.
In this case, the trial court properly applied a de novo standard of review to determine if the final agency decision applied the correct interpretation of the rules regarding whether a release had been discovered at One-Stop prior to 30 June 1988. We must now determine whether the trial court correctly applied that standard of review.
In order to be eligible to receive reimbursement for clean-up of leaking commercial UST’s, the discharge or release must have been discovered or reported after 30 June 1988. See N.C. Gen. Stat. § 143-215.94N. For purposes of determining whether a leak has been detected to establish eligibility to receive reimbursement from the Fund, a “ ‘[discovered release’ means a release which an owner or operator, or its employee or agent, has been made aware of, has been notified of, or has a reasonable basis for knowing has occurred.” 15A N.C.A.C. 2R0202(b)(4) (July 2003). NCDEHNR’s interpretation of this rule, as applied by the ALJ, the final agency decision, and the trial court, provides that a leak may be “discovered” either by analytical testing, official or unofficial notification, or through other factual circumstances. YOCO contends the appropriate standard should be that in the absence of specific knowledge of analytical testing results showing contamination, the only basis for detecting a leak should be upon official notification by NCDEHNR.
However, “an administrative agency’s interpretation of its own regulation should be accorded due deference unless it is plainly *555erroneous or inconsistent with the regulation.” Simonel v. N.C. School of the Arts, 119 N.C. App. 772, 775, 460 S.E.2d 194, 196 (1995). In this case, NCDEHNR’s interpretation of its own rule is neither plainly erroneous nor inconsistent with the regulation. NCDEHNR’s interpretation instead includes scenarios in which an owner or operator has specific actual knowledge of a leak, has been made aware of a leak, officially notified of a violation, or where there are sufficient circumstances that it is reasonable a leak should have been discovered, which are the exact scenarios that are encompassed in the broad language of the rule. Accordingly, the trial court did not err in affirming the interpretation of the regulation, applied to determine when a release was discovered, used in the Final Agency Decision and by the ALL
B.
In affirming the final agency decision adopting the recommended decision of the AU granting summary judgment to NCDEHNR, the trial court used the “whole record” test to determine that there was sufficient evidence to support the final agency decision that the release was discovered prior to 30 June 1988. Because the issue before the trial court was, however, whether summary judgment was properly granted, the correct standard of review remained de novo. Thus, the question before the trial court should have been whether there were any genuine issues of material fact and whether any party was entitled to judgment as a matter of law. See N.C. Gen. Stat. § 1A-1, Rule 56(c) (2003); see also N.C. Gen. Stat. § 150B-51(d) (scope and standard of review in reviewing a final agency decision allowing judgment on the pleadings or summary judgment).
In this case, the AU’s decision recommending summary judgment and the final agency decision contained a number of factual findings. This Court has previously discussed the role of findings of fact in a summary judgment order.
The entry of summary judgment presupposes that there are no issues of material fact; so findings of fact are not required. Nevertheless, it may be helpful in some cases for the trial court to summarize the undisputed facts which justify its order. If findings of fact are needed to resolve a material issue, however, summary judgment is improper and any such findings are disregarded on appeal. Accordingly, we must determine whether the . . . order is supported by the undisputed facts as they appear in the record without regard to the . . . findings of fact.
*556 Cieszko v. Clark, 92 N.C. App. 290, 292-93, 374 S.E.2d 456, 458 (1988) (citations omitted). In this case, the ultimate factual issue to be decided was whether YOCO had been “made aware of,... notified of, or ha[d] a reasonable basis for knowing” a release had occurred from its USTs prior to 30 June 1988. 15A N.C.A.C. 2P.0202(b)(4). This issue is one of material fact because a finding that YOCO had discovered a release prior to 30 June 1988 would make YOCO ineligible to receive reimbursement from the Fund. On the other hand, a determination that the release had not been discovered by YOCO prior to that date, would allow YOCO to be reimbursed from the Fund for clean up related to the leaking USTs.
NCDEHNR contends the grant of its motion for summary judgment was proper and points to the following facts in support of its argument that YOCO had a reasonable basis for knowing of the leaking USTs at One-Stop prior to 30 June 1988. One-Stop was located next to James’ property. The James’ property was located 150 feet down gradient from One-Stop; and One-Stop was the only gas station within a half-mile radius. Gasoline contamination of the James’ water supply was reported to NCDEHNR in 1986. An affidavit by Kay regarding his investigation showed that he interviewed One-Stop’s store manager who stated that One-Stop purchased bottled water for drinking and the odor of gasoline could occasionally be observed emanating from One-Stop’s water supply. Furthermore, Clark, One-Stop’s owner, stated that contamination had been noticeable in the water since shortly after a well had been installed in 1982. In September 1986, a letter from NCDEHNR informed Clark that One-Stop was suspected of petroleum contamination. Clark was also aware of a 1986 newspaper article discussing petroleum contamination of James’ water supply and notified YOCO of the article, and Gary York, the president of YOCO, admitted having been made aware of the contamination.
In considering whether summary judgment is appropriate, however, the evidence must be viewed in the light most favorable to the nonmoving party, see Dalton v. Camp, 353 N.C. 647, 651, 548 S.E.2d 704, 707 (2001), and “[a]ll inferences of fact must be drawn against the movant and in favor of the nonmovant,” Roumillat v. Simplistic Enterprises, Inc., 331 N.C. 57, 63, 414 S.E.2d 339, 342 (1992).
In that regard, YOCO presents conflicting evidence of record that shows the following in support of its contention that there was not a reasonable basis for discovering the leaking USTs at One-Stop. Once YOCO was made aware of contamination of James’ water supply, *557Collins Petroleum was hired to inspect One-Stop’s USTs. The Collins Petroleum testing found no leaks. The September 1986 letter to Clark from NCDEHNR showed only that One-Stop’s water had tested positive for fecal coliform bacteria, not petroleum contamination. A November 1986 analysis of One-Stop’s water, following the publication of the newspaper article, revealed “[n]o base/neutral or acid extractable organics detected.” Moreover, in May 1988, NCDEHNR sent a letter to James, which stated that at that point, NCDEHNR had not even made any determinations from its investigation, despite discovering petroleum in a monitoring well at One-Stop. NCDEHNR’s draft report was not issued until November 1988 and notices of violation did not issue to YOCO or One-Stop until 10 February 1989. Furthermore, there is no evidence that YOCO or One-Stop was made aware of the discovery of petroleum in the monitoring well in March 1988 prior to the issuance of the draft report or the notices of violation.
Viewing the evidence in the light most favorable to YOCO, we conclude that there is conflicting evidence on the issue of whether YOCO had a reasonable basis for discovering the leaking USTs prior to 30 June 1988, where even though the evidence shows YOCO was aware of petroleum contamination in the water supply of a neighboring property located down gradient from One-Stop, and that the odor of gasoline could occasionally be detected from One-Stop’s water supply: YOCO’s own testing by Collins Petroleum revealed no leaks from the USTs; NCDEHNR’s testing of One-Stop’s water supply revealed only fecal coliform bacteria contamination and “[n]o base/neutral or acid extractable organics detected”; and, NCDEHNR did not issue its notice of violation to YOCO until February 1989.
Therefore, there was a genuine issue of material fact to be decided as to whether a release had been “discovered” prior to 30 June 1988 within the meaning of 15A N.C.A.C. 2P.0202(b)(4) from the USTs at One-Stop. Thus, summary judgment in favor of NCDEHNR ruling that YOCO was ineligible to receive reimbursement from the Fund was improperly granted. Accordingly, we reverse the order affirming the final agency decision and remand this case to the trial court.
Reversed and remanded.
Judges WYNN and TYSON concur.