This appeal relates to a personal injury action concerning a single ear accident in which the passenger was killed. The deceased left behind a three-year-old son, and the provisional tutor filed suit on his behalf against the driver of the vehicle, her insurance company, and the State of Louisiana, through the Department of Transportation and Development (DOTD). The trial court found the DOTD liable and awarded damages in the amount of $575,910.00. We affirm and remand.
JaBASIC RECORD FACTS
On September 4, 1987, Deborah DiBene-detto was riding with Holly Sweeting on Airline Highway. They were traveling back to Baton Rouge, Louisiana, after leaving B.B.’s Lounge near LaPlace, Louisiana. As they approached the parish line separating St. John and St. James Parishes, the 1979 Datsun 280 ZX driven by Ms. Sweeting flipped out of control. Both women were thrown from the car. Holly Sweeting lost consciousness and suffered serious head injuries. Deborah DiBenedetto died as a result of her injuries.
The accident occurred about 2:30 a.m. as Ms. Sweeting was following another vehicle driven by Chris Picou. Ms. Sweeting was not cited by the state police trooper who investigated the accident. Tests revealed Ms. Sweeting’s blood alcohol level to be .05. Chris Picou testified Ms. Sweeting was following him in the left lane about two to three car lengths back as he drove in the right lane approximately 65-70 miles per hour. Although Holly Sweeting testified she has no recollection of the accident, Mr. Picou and his passenger, Rhonda LeBlanc, both testified they did not notice or remember running over any obstacles before the car wrecked.
ISSUES
In brief, the appellant specifies the following issues:
(1) Whether the trial court’s finding that DOTD violated its testing procedures is manifestly erroneous;
(2) Whether the trial court could lawfully rewrite DOTD’s uniformly-followed testing procedures;
(3) Whether the trial court’s finding that the highway was unreasonably dangerous is manifestly erroneous;
(4) Whether the trial court’s inferences that DOTD somehow created the “particular defect” which supposedly caused the accident are manifestly erroneous; and
| ,.¡(5) Whether the trial court’s inference that DOTD had notice of and an opportunity to remedy the “particular defect” which caused the accident is manifestly erroneous.
Appellee asserts the trial court erred in its limitation of the amount of damages awarded to the plaintiff was manifestly erroneous and, therefore, subject to reversal and remand for an increase in award under Chamberlain v. State, through DOTD, 624 So.2d 874 (La.1993).
ANALYSIS
Standard of Review
Our appellate review is what is commonly referred to as an Arceneaux review in that we must determine from the record whether there was a reasonable factual basis for the trial court’s factual findings and further determine whether the trial court erred as a matter of law. See Arceneaux v. Domingue, 365 So.2d 1330 (La.1978).
*388 ISSUES ONE AND TWO
The trial court’s Reasons for Judgment outline the three-step process needed to prove liability. First, there must be a finding that the thing which caused the damage was in the care or custody of the defendant; second, that it had a vice or defect which rendered it unreasonably dangerous; and third, that injury was caused by the defect. Sistler v. Liberty Mutual Ins. Co., 558 So.2d 1106 (La.1990).
The trial court noted the thing alleged to have caused plaintiffs damage is the highway, which includes the shoulders, and found it well-settled that the DOTD is vested with the responsibility of maintaining the highways of this State.
Secondly, the plaintiff argues there was a defect due to DOTD’s improper testing procedures of materials used in the reconstruction of the highway. This resulted in materials being used on the shoulders of the highway which did not meet the specifications set forth in the Gold Book, which is the Louisiana Standard Specifications for Roads and Bridges.
UThe trial court found the specifications of aggregate to be used on road shoulders is that it must have a passage rate of 100 percent through a two-inch sieve (Plaintiffs Exhibit 9B). DOTD argued it was its “policy” to round up the true passage rate of shells to the next whole number. Thus, the two samples, which were taken from the pile of aggregate which had a 99.6 percent rate of passage, did meet Gold Book specifications since it would have been rounded to the next whole number.1
We agree with the trial court’s interpretation of the Gold Book specifications. The Gold Book specifications reveal nothing regarding a rounding of percentages resulting from testing. Our review of the Gold Book reveals the specifications are usually set out in parameters. For instance, the gradation requirements for shell aggregate through a No. 4 sieve indicate 30-75 percent of the material must pass through the sieve. Another example is the specifications for foreign matter allowed in asphaltic surface aggregates (Plaintiffs Exhibit 9B, Section 1003.05). These parameters are set out in the form of highest limits. Nothing above this limit is tolerable. The logical conclusion is the Gold Book outlines parameters, which are the demarcation point of what the State will accept as usable material.
Regarding the material at issue, the Gold Book sets out the passage rate for shell aggregate through a two-inch sieve is 100 percent. There is not stated “zone of acceptance” like the other specifications, i.e., 95-100, etc. It is clearly stated that 100 percent must pass. Given the rest of the specifications in the Gold Book, we find that had a lesser amount been deemed acceptable, a lower acceptance parameter would have been established. Thus, the 99.6 percent of sample aggregate which passed through the two-inch sieve did not meet Gold Book’s specifications. Based on the ^evidence, we cannot say the trial court’s interpretation that 100 percent passage rate meant exactly 100 percent is clearly wrong.
Another issue that arose regarding whether DOTD violated its testing procedures was whether hand manipulation was permitted in the sieving process.
DOTD employees Hazel Hart and Nicholas Maekaluso testified it was common procedure to manually pass through any material that remained after the shaking process. Plaintiff argues this is clearly contradicted by the specifications (Plaintiffs Exhibit 20A). Such a procedure would allow the tester to manually manipulate materials until a 100 percent passage rate was achieved, thus never yielding a valid result. However, it is not clear from the record whether the remaining aggregate was manipulated (i.e., forced) through the sieve or merely dropped at a *389different angle than the horizontal shaking through the sieve.
The trial court found the Hammond DOTD office did not possess proper testing equipment, namely, a two-inch sieve. However, this is an error. The Hammond office did not possess a two-inch sieve to go with the mechanical shaker. It did possess a two-inch sieve in the manual shaker. Ms. Hart testified DOTD’s testing procedure consisted of using the mechanical shaker with a one and one-half-ineh sieve, and then placing any material not passing into the manual shaker into the two-inch hand-held sieve and shaken. Thus, the trial court’s statement that the Hammond office did not possess a two-inch sieve is error since the DOTD did have a hand-held sieve.
Additionally, we do not find the trial court was manifestly erroneous in finding the shells used in the inside shoulders of Highway 61 (used not in reconstruction, but in spot repairs) were never tested. Maurice Jorden, DOTD employee, testified no test was performed since only 500 cubic yards were needed for this section of repairs. The trial court interpreted the specifications that allowed visual inspection for maintenance requiring under 500 cubic yards referred to stone only (Exhibit 18B). We agree this does not relieve DOTD of the responsibility of sampling the stockpile I sfrom which material is taken. Here, no one was able to say whether the materials used met the Gold Book specifications because no evidence of any test being performed was found. Based on the foregoing, we cannot say the trial court was clearly wrong in finding DOTD violated its own testing procedures thereby causing the material to be defective.
DOTD argues that LDHTR 113-75 (Plaintiffs Exhibit 20) allows the result of the sieve analysis to be reported by rounding the percentages to the nearest whole number. DOTD employees Hazel Hart and Sarah Kemp both testified this rule was proper and uniformly followed. DOTD further asserts that the two engineers called by the plaintiff testified there was nothing in DOTD’s policies which prohibited rounding. In response to this argument, plaintiff points to the testimony of DOTD employees and contractors who all interpreted 100 percent to mean 100 percent and nothing less.
DOTD relies on La. R.S. 49:963 for the authority the trial court cannot lawfully rewrite DOTD’s uniformly followed testing procedures. The statute reads:
A. (1) The validity or applicability of a rule may be determined in an action for declaratory judgment in the district court of the parish in which the agency is located.
(2) The agency shall be made a party to the action.
B. (1) If, before the date set for hearing, application is made to the court for leave to present additional evidence, and it is shown to the satisfaction of the court that the additional evidence is material and that there were good reasons for failure to present it in the proceeding before the agency, the court may order that the additional evidence be taken before the agency upon conditions determined by the court.
(2) The agency may modify its findings and decision by reason of the additional evidence and shall file that evidence and any modifications, new findings, or decisions with the reviewing court.
C. The court shall declare the rule invalid or inapplicable if it finds that it violates constitutional provisions or exceeds the statutory authority of the agency or was adopted without substantial compliance with required rulemaking procedures.
D. An action for a declaratory judgment under this Section may be brought only after the plaintiff has requested the agency to pass upon the validity or applicability of the rule in question and only upon a showing that review of the validity and applicability of the rule in conjunction J¿with review of a final agency decision in a contested adjudicated case would not provide an adequate remedy and would inflict irreparable injury.
Our interpretation leads us to conclude the statute does not apply because it clearly states it is to be used in an action for declaratory judgment in order to determine the validity or applicability of a rule of a state agency. This is not a declaratory judgment *390action. Ergo, we find the trial court did not illegally rewrite DOTD’s testing procedure, but merely held DOTD to abide by its own published standards and specifications.
ISSUE THREE
Appellant asserts the trial court committed manifest error in finding the highway was unreasonably dangerous. A review of the trial court’s Reasons for Judgment reveals:
Additionally, there was uncontroverted testimony that the shells used on the inside shoulders were never tested. Maurice Jerden [sic], a DOTD employee, testified that no test was performed because only 500 cubic yards were needed. The parties did not agree that only a visual inspection was required. It is plaintiffs position that the specification allowing visual inspection for maintenance requiring under 500 cubic yards, refer to stone only. This is also the court’s interpretation of the specification. (See plaintiffs exhibit 18b). Further, it is also the opinion of the court that this does not relieve DOTD of the responsibility for sampling the stock pile from which the material is taken. In this case, there is absolutely no record of any tests being performed, not even the visual inspection. Therefore, Jerden [sic] was unable to say whether the materials used satisfied the Gold Book specifications. The evidence strongly suggests that much of the materials did not.
This court is of the opinion that the manner in which DOTD carried out its responsibilities, as per Gold Book specifications, leaves much to be desired. Several DOTD employees testified that they were familiar with the Gold Book, yet things were done the Vay they always did it,’ often in direct contradiction to the specifications. Thus, an unreasonably dangerous condition was permitted to exist on that portion of Highway 61.
Apparently, the trial court concluded that the very existence of a vice or a defect automatically created an unreasonably dangerous condition. We disagree. The Isproper analysis is to examine three relevant factors, which include the likelihood the defect will cause harm; the severity of the harm; and the cost of avoidance. Entrevia v. Hood, 427 So.2d 1146 (La.1983). Since we are vested with the authority to review fact and law (see La. Const, art. V, § 10), we will now determine if an unreasonably dangerous condition existed.
Likelihood of the Harm: DOTD did not use proper testing procedures and therefore the materials which did not meet its specifications were allowed to be placed on the shoulders of Highway 61. This was not a temporary condition but rather DOTD placed this material there as part of the permanent reconstruction of this section of Highway 61. Prior cases have held that the shoulder of the road is supposed to allow a motorist who accidentally or temporarily leaves the black top to maneuver his car back onto the highway. By using untested material and material not meeting the specifications in the Gold Book, the aggregate placed there may not have the capability of “settling” to form such a shoulder; thus, the likelihood a motorist may encounter trouble when going onto a shoulder is great. Also, by using material not conforming to its own set specifications, the shoulder would be filled with materials bigger than anticipated, which could create a hazard for a motorist’s tires.
Severity of the Harm: We presume there is a reason why the Gold Book’s specifications were set. They represent the safest and most useful specifications to allow roads and bridges to be functional. By using material not meeting the 100 percent two-inch sieve pass rate, it arguably puts bigger pieces of aggregate on the highway shoulder, which can be kicked into the roadway causing a tire to puncture or, if on a shoulder, causes the aggregate not to be as cohesive, thus allowing a car to fling shells onto the road or get stuck. The severity of this harm could range from a waste of time to help a motorist get his stranded car off the shells or a blow-out which, as seen, can cause a car to lose control and injure passengers or other passing motorists.
IsCost of Avoidance: DOTD can and should abide by its own published policy of testing materials. DOTD has set up what it deems acceptable material, merely following that standard is simple enough. It does not *391serve any social utility to have DOTD not follow its own rules. The rules themselves are set up to protect the public and allow for safe highways. Ergo, we find DOTD’s failure to follow its own proper testing procedures and placement of defective materials on the highway shoulders caused an unreasonably dangerous condition to exist.2
ISSUE FOUR
The trial court made a factual finding that it was probably an oyster shell which punctured the left front tire of Ms. Sweet-ing’s vehicle, causing the driver to lose control of the car. Our review of the record reveals one tire expert, Bobby Smith, testified there was no evidence of poor maintenance of the tire. It was his opinion the left front tire was penetrated by a large jagged irregular-shaped object, like an oyster shell. Mr. Smith even viewed the scene on September 18, 1987, soon after the accident, where he collected large oyster shells (shown not to be able to pass through a two-inch sieve) in the vicinity of the accident.
Another tire expert, Mark Laferriere, joined in Mr. Smith’s opinion that it was a large jagged object, like an oyster shell, that had caused the blow out. However, on I locross-examination, Mr. Laferriere admitted it was not possible to tell exactly what punctured the tire.
Direct testimony of Holly Sweeting, the driver of the vehicle, reveals she was not cited by the trooper who investigated the accident.
The trial court recognized the cause of an accident proven by circumstantial evidence need not negate all other possible causes. See Henderson v. Diamond Datsun, Inc., 413 So.2d 542 (La.App. 4th Cir.1982). Given the burden of proof is more probable than not, coupled with the prior proof that DOTD used shells not meeting the specifications, we cannot say the trial court was clearly wrong in finding the DOTD created the defect which caused the accident.
ISSUE FIVE
Appellant finally asserts the trial court was manifestly erroneous in its inference that DOTD had notice of and opportunity to remedy the .“particular defect” which caused the accident.
La.R.S. 9:2800(B) provides that, in an action against a public entity, based upon La. Civ. Code art. 2317, plaintiff must prove the entity had actual or constructive notice of the defect to warrant liability. La.R.S. 9:2800(0 defines constructive notice as the existence of facts which infer actual knowledge. The record clearly reveals the DOTD workers who testified and tested the shells knew they did not meet the 100 percent passing requirement, but rather only 99.6 percent passed. It also reveals the DOTD employees did not use enough samples to test the material and that the repairing materials eventually placed on the inside shoulder were never tested by DOTD. Because it is the responsibility of DOTD to maintain the highways and shoulders and DOTD knew it did not follow the proper procedures, we find there was constructive knowledge of the defect. The record shows the material was placed in the area in early May 1987 and the accident occurred in September 1987, some four *392I nmonths later.3 Thus, there was ample opportunity to repair the defect. We find the trial court’s finding was not clearly wrong. ISSUE SIX (BY APPELLEE)
Appellee raises the issue that the trial court’s limitation of the amount of damages awarded was manifestly erroneous and subject to reversal and remand for an increase under Chamberlain v. DOTD, supra.
In Chamberlain, the Louisiana Supreme Court held La.R.S. 13:5106(B)(1), which would impose a $500,000 ceiling on general damages recoverable in a personal injury suit against the State of Louisiana, its agencies, or its subdivisions, was unconstitutional. The court held that the ceiling contravened the constitutional proscription against sovereign immunity contained in La. Const, art. XII, § 10. In light of the declaration by the supreme court, we remand the matter to the trial court so that an increased damage award may be issued in light of the uneonstitutionality of La.R.S. 13:5106(B)(1).
For the reasons assigned, the trial court’s judgment is hereby affirmed and the matter is remanded, as per the views expressed herein. All costs of this appeal are to be assessed against the appellant.
AFFIRMED AND REMANDED.
. It was also established the number of samples taken from the aggregate should have been three, instead of two, since it was DOTD policy to get a sample from every 1,000 cubic yards to be used. If under 1,000 yards, one sample would be taken, if greater than 1,000 but less than 2,000, two samples would be taken. If it was greater than 2,000 but less than 3,000, three samples were to be taken. The work order revealed 2,362.5 cubic yards were ordered for the project; thus, DOTD should have tested another sample.
. Research has revealed that in cases where the DOTD was held liable there are some common factors. One of these factors is that there be a substantial quantity of shoulder material on the hard portion of the road. There must also be a showing that the DOTD was negligent in not scraping this material off. There is also the factor that eyewitnesses on the scene soon after the accident have all testified that they, too, saw the condition of the highway and that there was no reconstruction going on when the accident occurred but the condition was more of a permanent nature. These things are reflected in the following cases: Llorence v. State, DOTD, 558 So.2d 320 (La.App. 3d Cir.1990), writs denied, 565 So.2d 443 (La.1990); Schexnayder v. State, 477 So.2d 1175 (La.App. 1st Cir.1985) and Bennett v. Nat’l Fire & Marine Ins. Co., 472 So.2d 256 (La.App. 1st Cir.1985), wrif denied, 476 So.2d 352 (La.1985). We note that the DOTD relies heavily on the case of Duffy v. State, through DOTD, 415 So.2d 375 (La.App. 1st Cir. 1982), writ denied, 420 So.2d 448 (La.1982) for the proposition that oyster shells are not unreasonably dangerous per se. However, we note that that case is distinguishable on the facts because there the accident occurred during grading operations and the court held that the utility of having the state repair its roads did not outweigh the risk that a motorist passing through a known construction site would encounter repair materials on the road.
. The record reveals the DOTD "accepted” the completed project in August 1987, approximately two weeks prior to the accident.