Review of Improving soil aquifer treatment efficiency using air injection into the subsurface

Ido Arad1, Aviya Ziner1 Shany Ben Moshe1, Noam Weisbrod2, and Alex Furman1

The presented study investigated the effect of air injection on infiltration rate, biogeochmical processes and effluent quality in long (200 cm) column experiments that simulated the recharge and treatment of secondary treated wastewater in a soil-aquifer treatment (SAT) system. Different air injection frequencies and durations were tested and compared to the normal operation of SATs, which consists of well defined wetting-drying cycles. The manuscript is well written and clearly presents the experimental setup and results of the experiments. Tradeoffs in pollutant removal vs. infiltration rate are discussed for major constituents such as nitrogen species, DOC, and contaminants of emergent concern (e.g. ibuprofen). The results clearly indicate that air injection can increase the wastewater volume that can be recharged and treated with SAT if air is injected into the subsurface as the injected air breaks down crusts, biofilms and creates new preferential flowpaths. The study is clearly of interest to the hydrologic science community as groundwater recharge is growing in demand worldwide. I recommend publication of this paper after minor revisions.

Specific comments

Line 18, abstract: Please add the length of the columns in the column description (e.g. after six long column…)

Line 24: “doubled the infiltration time” might be misleading as wording. Suggest changing to “double the time during which infiltration was possible”

Line 27: What is SLPM?

Lines 59-62: Could you please state threshold values or value ranges for oxygen for some of the processes listed here? What do you consider aerobic conditions? Under what oxygen conditions does the ammonification rate decrease substantially? Is there a threshold that can be defined as clear tipping point? Does nitrification assume the same oxygen conditions (same number ranges)?

Line 72: What is the mechanisms behind the higher removal of CECs during longer drying periods? Is it just the aerobic conditions?

Line 82: The column description is a bit confusing. Was the column only 200 cm long with the top 30 and the bottom 30 cm kept empty to provide room for flooding and drainage. That would suggest only 140 cm were effectively filled with soil? Also what are the 20x10x10 modules? How were they connected? How did you pack the column and avoided capillary boundaries between the layers/modules?

Line 108: By “surface pressure head” you mean the hydrostatic pressure of the ponded water?

Line 158: SLPM still not defined …but finally given in Table caption for Table 1.

Line 196: Why calculate a mean ORP for the profile if there is clearly a gradient with depth?

Line 199: Why did you choose to use glucose, which is a very digestible form of carbon? What forms of carbon are typically found in the wastewater?

Line 283: The fast developing clogging effect is really striking! Have you thought about adding soil microbial analysis to investigate whether the abundance of bacteria is increasing supporting the idea of biofilm formation?

Line 385: These results suggest that the air perhaps should be injected even deeper (>1m) in the profile? Air buoyancy will ensure that air rises to the top of the profile but perhaps the deeper injection can address some of the low oxygen concentrations in the deeper profile.

Line 465: It seems maintaining aerobic conditions in the upper profile provides suitable conditions for mineralization and nitrification and having low oxygen and reduced conditions in the deeper profile creates the right environment for denitrification. Often natural soils show a huge decrease in microbial abundance from the top 10 cm of the soil profile to 1 m. What is the microbial abundance on denitrifiers in the deeper soil profile that can actually reduce large amounts of nitrate?

The paper presents experiments on the impact of air injection into a soil column with infiltration of water from top to bottom. The background of the study is soil aquifer treatment as an efficient waste water treatment method. The goal is to enhance biodegradation of substances in the effluent. As important degradation processes are aerobic, aeration is necessary. Common practice is to alternate infiltration and drying periods. The paper provides a proof of concept study for air injection in the subsurface with steady infiltration as an alternative method that allows for larger volumes of effluent to be treated in a given time span. For this purpose, infiltration and air injection with different rates and durations have been tested with the column and water content and water quality have been monitored. Concentrations of different substances were used to quantify water quality. The finding of the paper is that air injection can lead to improved water quality and higher effluent rates that can be treated, however, the air injection cycles have to be adapted. The results depended on the injection periods and rates. 

The paper is clear and well written. The findings are interesting and relevant for HESS. I have only minor comments that concern mostly clarifications. The only larger comment is about formulations. Some of the interpretations of the findings are formulated as results, which is not really covered by the experimental observations. They are interpretations and this should be made a bit clearer. 

•    The manuscript uses many abbreviations. This is a question of taste, of course, but I had to scroll back and forth a lot during reading. A table might be useful.

•    Figure 1 and Section 2.2: It did not become clear to me how the infiltration was carried out. I might have missed the description. From the text I understood that the infiltration was done by a constant head that was realized with the overflow tank. The resulting infiltration rates were not constant and depended on the soil conditions. In Figure 1, there is a peristaltic pump that seems to inject water with a constant rate. If water was pumped with constant rate, I would not understand how this could be, as the infiltration rate was measured and was not constant. 

•    Line 145-146: I did not understand the sentence. What is meant by no pressure head above the unsaturated soil surface? Does this mean no ponding water? If there is water, the water always has a pressure head, which can be zero or positive or negative, but there cannot be no pressure head. I guess what is meant is no water. 

•    Line 146: I also did not understand how the effluent supply can be higher than the infiltration. From figure 1 I understand that the effluent is the infiltrated water. Or were there separate supplies of effluent and clear water?

•    Line 191: The ‘h’ is missing after the times. 

•    Lines 192-194 and Figure 2: I was here also confused about the infiltration rates. If I understand correctly, the infiltration rates were concluded from the volume in the tank. How often was this done? In Figure 2, the mean infiltration flux is given and I assume that this is the time average of the infiltration flux. At least this would explain why it is so constant over longer time spans in figure 2. What time intervals were chosen for averaging? And why? Why not showing the time series of the infiltration rate without averaging? This would allow to get a better picture of how much the infiltration rate is related to the water content in the upper soil column part. 

•    Related to the previous comment: If I understood the averaging of the infiltration correctly, I think it would be better to use the term ‘mean infiltration’ consistently in the text instead of calling it ‘infiltration’ (example lines 275, 279). This would prevent misunderstanding.

•    Section 3: Do I understand it right that the first 24 hours of the experiments FDO and AI-LF1 were the same? The large difference of water content in the upper soil at the beginning is thus the inherent uncertainty in the experimental results. I think this has to be considered when conditions in the experiments are compared. Smaller deviations have to be interpreted a bit carefully. 

•    Lines 259-266: This paragraph contains an interpretation of the increasing water content and infiltration rates with time in the air injection experiments with low infiltration rates. The interpretations are plausible, however, they are just interpretations. There are other possibilities that could explain the findings, such as redistribution of water with the air injection that leads to connected patterns with higher water saturation and thus higher conductivity, facilitating higher infiltration. The increased water content could be local, and would then be rather accidentally. Also, the interpretation with the breaking of crusts reads a bit odd. I am not an expert on bio-clogging, but the time seems very fast to me. If breaking of clogging crusts was relevant for the increase of injection, the crusts must have formed before the first 24 hours. Is it realistic that this happens so fast? Is there evidence for crust building in the sand column (maybe visible at the surface)? I am also not so convinced by the preferential pathways. Why should a re-arrangement of grains (this is how I would interpret the ‘creating wider pores’ read) should lead to large pore clusters connected from top to bottom? It is possible, but not very likely. To clarify: The interpretations are plausible and I would not argue against them. But the preferential pathways and crusts are later in the manuscript treated as results. I think they need to be kept as possible processes that take place, not as given ones. 

•    Line 278: If the water content in the whole column decreased, where did the water go? Could one observe a larger outflow from the column that matches the decreased water content?

•    Line 350-351: This conclusion is here a bit misplaced and should better be moved to the end, when also the water quality was discussed.

•    3.2: ‘Effluent quality’, not ‘effluent quality’

•    A general comment: Would it for practical application not be important to study the range of influence of an air injection point? In the experiments, air influenced the water quality in the whole column, but as is later written in the discussion, will probably have an influence only to a certain depth. The column had a not too big cross section, so that the air flow was directed mainly vertically. In a real field a large area would have to be reached. This might involve a lot of injection points. 

The point is not that this should be answered in the paper, it is just something one thinks of directly after reading the conclusions.