31. Reasonableness Review-Sandhill

Sandhill is a water treatment plant that began operation in 1965 to treat water from Lytle Creek. By 1969, it was capable of treating up to 20 million gallons of water per day (mgd). However, it was not approved for treating water from the State Water Project (SWP).

During storms, periods of heavy snow melt and periods of high recreational use, Lytle Creek water has high turbidity.²⁴ Sandhill's limited ability to handle high turbidity caused it to have to be shut down during periods of high turbidity because it could not meet federal and state turbidity requirements.

After subsequent upgrades, Sandhill was authorized to treat SWP water in 2002. However, it was required to blend Lytle Creek water with SWP water such that the mixture contained at least 20% Lytle Creek water. As a result, when Sandhill was unable to supply sufficient Lytle Creek water, the SWP water could not be used.

In A.05-08-021, SGV proposed to upgrade Sandhill and expand capacity to 29 mgd at a cost of $35 million. The planned upgrades would allow Sandhill to treat 100% Lytle Creek water, 100% SWP water or any combination of the two. The need for the upgrade, including cost-effectiveness, was addressed in D.07-04-046. In Section IX.B.1(c), of that decision, the Commission stated:

"We find the Sandhill treatment facility to be needed and building it is reasonable."

In Finding of Fact 42, the Commission stated:

"The Sandhill plant is cost-effective and it is reasonable to construct it."

Thus, the Commission has found that the Sandhill project is needed.

31.1. Positions of Parties

DRA states the construction contract was not awarded to the lowest bidder and SGV has been unable to provide any documents that demonstrate that the lowest bidder was not qualified as it claims. For this reason, DRA recommends the removal of $4,944,000 from total costs.

DRA recommends a disallowance of 33.9% of any otherwise reasonable costs from ratebase because only 66.1% of Sandhill capacity will be used and useful. DRA bases this recommendation in part on its determination that the modifications will only be useful to treat Lytle Creek water for turbidity for 37 days per year more than would have been the case without the modifications. The other basis for its recommendation is a December 15, 2008, press release by the California Department of Water Resources that SWP water deliveries could be permanently reduced by 50%. DRA also recommends that SGV be allowed to include more of the costs in ratebase in future GRCs if it shows that it will have sufficient water supplies to utilize more Sandhill capacity.

COF states the actual costs of Sandhill are over $40 million and approaching $60 million if related projects are included. COF argues that Sandhill has never operated at 29 mgd and is physically incapable of doing so as currently constructed. COF also argues that Lytle Creek water is insufficient to provide reliable and consistent supply to make the 29 mgd capacity used and useful.

COF alleges the pretreatment facility has a capacity of 20 mgd.

COF states that water demands, when surface water is available, are considerably below plant capacity and only limited water will be available when demand increases.

COF states that costs for production of treated water are greater than for other sources. COF states that use of SWP water to supplement Lytle Creek water is not economical and the amount available is limited and should not be considered a reliable source.

COF argues that the construction costs are excessive because SGV did not use competitive bidding and the costs are 40% higher per gallon per day than other recently constructed adjacent filtration plants.

COF argues costs in excess of the $35 million cap should automatically be disallowed as violating the cap.

For the above reasons, COF recommends a maximum of $17,000,000 of the costs be allowed in ratebase. If Sandhill capacity and demand increases in the future, COF recommends that an additional portion of the costs be considered for inclusion in ratebase in future GRCs if SGV demonstrates that the costs are reasonable.

FUSD states that there are limitations on the amount of water SGV can take from Lytle Creek and the SWP, and other water sources would not be cost-effective. FUSD also represents the maximum amount of water that can flow to Sandhill through the pipeline that provides the water is 18.7 mgd, which is close to the capacity of Sandhill without the upgrade. FUSD points out that the pretreatment facility has a capacity of 20 mgd. Therefore, the additional capacity is not used and useful.

FUSD argues that the construction costs are excessive because SGV did not use competitive bidding. For example, the costs are as much as 57% higher per gallon per day than a recently constructed upgrade at another nearby filtration plant built by the same contractor.

FUSD recommends that all costs associated with the expansion of Sandhill capacity be excluded from ratebase. FUSD estimates the costs to be removed as $16.7 million for the expansion plus $4,944,000 due to the lack of competitive bids, for a total of $21,644,000.

SGV argues that its selection of R C Foster (Foster) as general manager was reasonable, the resulting costs were reasonable and the upgrades are used and useful.

31.2. Discussion

As discussed above, the Commission determined in D.07-04-046 that the Sandhill upgrade is needed, and cost-effective. Thus, questions related to the need for the project and its cost-effectiveness are not appropriate for this proceeding and will not be addressed herein.

There are pipeline projects that are related in part to the Sandhill upgrade because they are or will be capable of transporting water from Sandhill. They were not subject to the $35 million cap on Sandhill costs and are addressed separately in this decision on their merits.²⁵

The questions raised by the parties address:

1. The reasonableness of the costs;

2. The amount of Lytle Creek water that is available to be treated by the plant; and

3. How much Lytle Creek water the plant is physically capable of treating.

These issues will be addressed in sequence.

The parties recommend disallowances of certain costs because Foster was not selected by competitive bid.

SGV explains that it was aware that Black & Veach Corporation (B&V) had recently designed and managed construction of a similar plant for West Valley Water District (WVWD) to treat both Lytle Creek and SWP water. B&V had also designed a water treatment facility for Cucamonga Valley Water District (CVWD) to treat SWP water. SGV discussed B&V's performance with WVWD and CVWD. Based on these discussions and B&Vs overall experience, SGV chose B&V to design the Sandhill modifications.

SGV had B&V obtain proposals from experienced contractors to do the construction with B&V functioning as the general contractor. SGV ultimately decided not to use B&V as the general contractor because it and B&V could not reach agreement on contract terms and conditions. SGV then sent a request for proposal to Foster with whom SGV had a track record of successful water treatment projects. Foster's bid for construction was slightly higher than the bid submitted by B&V, and Foster accepted the contract elements B&V rejected.²⁶ SGV then awarded the construction contract to Foster.

SGV's selection process for general contractor initially focused on B&V because of its successful completion of very similar nearby projects. When it could not reach agreement with B&V, SGV went to Foster because of its successful experience with Foster. Ultimately, Foster's bid was slightly higher than B&V's. While this process is not strictly a competitive bid process, it did result in consideration of comparable bids from two well qualified contractors. While not a perfect process, it appears reasonable. Thus, there is no reason to believe the choice of the general contractor without a competitive bid process resulted in unreasonable costs.

In D.07-04-046, the Commission imposed a cap on rate base additions of 10%, but exempted Sandhill from the cap. The cap for Sandhill was set separately at $35 million. The cap was set to limit annual additions to plant. In the case of Sandhill, the amount in ratebase was to be increased by advice letter each year and the cap provided an overall limit for the GRC cycle. The decision does not indicate that the cap was intended as a permanent overall cap on the Sandhill upgrades.

The next issue between the parties is whether there will be sufficient water available to run Sandhill at a full capacity of 29 mgd. This is an issue related to the need for the upgrades, which was addressed in D.07-04-046. However, it will be discussed herein.

Lytle Creek water comes to Sandhill from the Southern California Edison Company (SCE) afterbay below its power house on Lytle Creek. The afterbay is fed by a penstock from SCE's powerhouse. The penstock is a 36-inch diameter pipe. SGV represents the penstock was designed to handle up to 40 mgd to the afterbay. In addition, SCE has a contractual obligation to deliver up to 38 mgd to the afterbay.²⁷ From the afterbay, 6.7 mgd is diverted to other water providers.

DRA, COF and FUSD have argued that SCE cannot provide sufficient water to the afterbay for Sandhill to operate at its design capacity of 29 mgd because historical data on flows through the penstock did not indicate it could do so. However, the record shows that stream flows in Lytle Creek have been sufficient to do so on some occasions, but SCE has not diverted all of the water it could to the penstock. Overall, the record shows that 29 mgd can be made available in the afterbay for SGV's use when the water in Lytle Creek is sufficient.

DRA, COF and FUSD have argued that Sandhill is not physically capable of treating 29 mgd. Specifically, they argue that 29 mgd cannot be brought to the plant through the 36-inch pipe from the afterbay. They also argue that the pretreatment facilities within the plant have a capacity of only 20 mgd.

The 36-inch diameter pipe from the afterbay to Sandhill is gravity fed which means that there are no pumps to increase the flow. Therefore, the amount of flow, assuming water is available at the afterbay, is a function of the available head. The available head is the difference in elevation between the surface of the water in the afterbay and the surface of the water where it is delivered to Sandhill. Working against the flow in the pipe is the headloss. This is the effect of the amount of resistance to flow in the pipe system expressed in equivalent feet of head. Head loss is caused by the roughness of the inside surface of the pipe. Other sources of head loss include any place where the direction of the pipe flow changes such as a bend, any fittings such as a valve or meter, changes in diameter, the shape of the inlet to the pipe and the shape of the outlet from the pipe. The amount of the head loss increases as the flow rate increases. The maximum possible flow in the pipe is achieved when the available head equals the headloss. All of the parties provided calculations of the available head and the head loss. The record shows general agreement among the parties on most of the inputs to the calculation of head loss. However, there is disagreement as to the available head.

COF and FUSD measured the available head as the difference in elevation between the surface of the water in the afterbay and the surface of the water in the influent equalization reservoir. This provides an available head of 3.77 feet. With this amount of available head, only about 22 mgd could be provided to the pretreatment facilities. However, since the pretreatment facilities can only treat 20 mgd, COF and FUSD assert the plant's capacity would be limited to 20 mgd.

However, the influent equalization reservoir is only needed to blend Lytle Creek water with other sources such as SWP water. Water from the afterbay can be directly fed to the pretreatment facilities resulting in an available head of over 10 feet, which is more than sufficient to allow 20 mgd to the pretreatment facilities. At the same time, nine mgd of Lytle Creek water can be diverted toward the original diatomaceous earth filters. This water would be mixed with up to eight mgd of water that has been pretreated. This would reduce the turbidity of the mixture allowing it to be treated in the diatomaceous earth filters. The resulting 17 mgd of water from the diatomaceous filters would then be mixed with the 12 mgd of pretreated water not sent to the diatomaceous filters. The result would be 29 mgd. Therefore, Sandhill is physically capable of treating 29 mgd of Lytle Creek water if sufficient water is available.

As discussed above, the Commission determined in D.07-04-046, the upgrade of Sandhill was reasonable. The upgraded Sandhill water treatment plant is in service providing water to SGV. On those occasions when the flows in Lytle Creek are sufficient, up to 29 mgd can be made available to SGV in the SCE afterbay. Sandhill is capable of treating up to 29 mgd of Lytle Creek water when available. Thus, Sandhill is used and useful and its costs are reasonable.

²⁴ Turbidity is the amount of solids, in this case primarily China clay, suspended in the water.

²⁵ See Sections 11, 16 and 23.

²⁶ Foster's bid was for construction to be done by a joint venture of it and Coletta Corporation.

²⁷ As a practical matter, since Sandhill could not previously handle more than 17 mgd from the afterbay, there has been no need for SCE to supply the greater amount.