Background of the project

When considering a visit to the wine country most people think of tasting some special wines, picnicking in a lovely setting and perhaps staying a night or two at a quaint bed and breakfast. Most people do not think of touring the wastewater management facilities. Indeed, most people do not think of wastewater at all. But I do. I try not to think of it as waste, but rather, like a weed in a garden whose value has yet to be fully realized.

Many people have fairly negative feelings when they consider wastewater, especially the municipal variety, yet treated wastewater can become "cleaner" than some surface waters. And wastewater management facilities do not need to tucked into an out-of-the-way corner, or hidden from public view. Ideally, a wastewater management system can become part of the landscape, providing a valuable resource the surrounding landscape.

Lately (in the last 5 years) I, with the help of friends and colleagues at the University of California at Davis (U.C. Davis), and a couple of wineries, have been exploring the potential for constructed wetlands to treat the process wastewater generated from winery operations. The largest portion of the wastewater is produced during the crush season, roughly from late August through October, when the harvested grapes are crushed to release their juice.

The bulk of the remainder of the wastewater is produced during racking, when the wine is removed from the primary fermentation tanks and the tanks are washed down, and during bottling process when steam is frequently used for sterilization. Depending on the size of the winery in question, bottling can take place during a single month, or can occur continuously throughout the year. As may be notable thus far, the production of wastewater in wineries is not continuous but rather varies considerably in terms of both quantity and quality. This state of affairs presents a challenge in wastewater treatment, one that constructed wetlands may be able to help solve.

Design of the wetland

For three years (1996-1998) a pilot subsurface horizontal flow constructed wetland (CW) system was operated at U.C. Davis. A subsurface flow design was chosen in order to control potential odors (which we had learned from experience could be substantial) and to eliminate potential mosquito problems. The idea was to determine just how much of an organic load the CW could tolerate without 1) "failure" (defined by clogging, failure to remove organic load, or producing objectionable odors) or 2) "death"(defined as visible extensive damage to the plants).

The CW was filled to a 1m depth with washed pea gravel and planted with bulrush (Scirpus acutus) and Cattail (Typha, spp). The CW was operated as a batch feed system in that the wastewater entered the holding tank once per day, but functionally as a continuous feed system in that it took approximately 22 hours for the tank to empty while losing less than 10% of the flow rate. The retention time was initially 11 days and dropped to 9.7 days after 3 years (as determined by tracer studies).

Results

We knew from the beginning that constructed wetlands could do a great job of removing organic constituents, but we did not know how great. We guessed that, because of the quality of winery wastewater (which contains sugars, short-chained organic acids, some proteins, and some phenolic compounds as compared to municipal wastewater which contains appreciable oil and fats), our system would perform better, in terms of bulk organic removal, than those used for municipal wastewater, and we were right.

Generally speaking, it is advised that constructed wetlands not be loaded at an organic loading rate (OLR) greater than 110 kg BOD₅/ha^.d (about 200 kg COD/ha^.d). This rate is advised to limit odors and prevent clogging. We used OLRs ranging from 345 kg COD/ha^.d to 1640 kg COD/ha^.d, up to 8 times the recommended loading rates, to evaluate performance under high organic loads. In addition, the loading rates were varied both daily and seasonally to evaluate performance under variable conditions such as those experienced at wineries. Of particular note is that all of the wastewater experienced pretreatment such that the settleable solids load was less than 2% at all times.

Better than anticipated

The results were far better than anticipated. Summarized, the experiments resulted in treated effluent containing less than 100 mg/L COD (about 40 mg/L BOD₅- note that the COD/BOD ratio changes as treatment progresses) 95% of the time (on only 3 occasions did the COD exceed this number). This level of treatment is significant in that in order to be used for irrigation (land disposal), the BOD₅ concentration may not exceed 50 mg/L. Clearly, with minimal effort, the system could be reliable in producing water of a quality acceptable for landscape irrigation. The clarity changes in the wastewater are illustrated in Figure 1: the flask on the left contains the original wastewater, that in the center is a sample taken from midway, and the flask on the right contains the final effluent.

Under pilot scale operating conditions, the CW was certainly able to manage variable and high strength wastewater without failure: The organic load was treated, the system did not clog and objectionable odors were not observed. The plants did not suffer dramatically — only at very high loads (an accidental overloading of 3500 kg COD/ha^.d lasting 6 days) was plant die back observed and then only in the first 0.6 m of the CW tank. But what about performance in the field?

We have had the chance to investigate this and have determined that yes, constructed wetlands can work very well for real wineries in the field but that solids separation is critical to maintained performance. Our first full-scale system (0.45 hectares) was removing up to 2500 kg COD/ha^.d in the first season of operation! This removal rate is even greater than that observed in the pilot system. However, in the second season of operation severe clogging was observed attributable to both very high ORL (up to 3600 kg COD/ha^.d) and to failure in solids separation.

Figure 1: Wastewater samples at three locations along the Constructed Wetland Bed

Restrictions of field trials

Field trials are often subject to user/owner restrictions, and the inability to control operations. Wineries are run by human beings -often at late hours (during crush)! While the entry of solids (pomace and lees) to the wastewater management system should not occur routinely in wineries, human failure does result in occasional spills. Without a failsafe solids separation approach these spills can be lethal to a constructed wetland. The lesson learned here???? No solids allowed. Ever.

A second constructed wetland has been completed for use at a winery and has just seen its first season of operation. Results aren’t yet available, but the system appears to be just fine. The solids separation screen is operating well, and the final effluent contains less than 10 mg/L COD (~5 mg BOD₅/L)!!!

Now entering its third season of operation, the first CW system has experienced an overhaul: The solids removal system has been improved and the CW system thoroughly cleaned. We await and anticipate even better results in the upcoming season and will keep you informed!

--> To table of contents